AAST 2013 PLENARY PAPER
Hemostatic resuscitation is neither hemostatic nor resuscitative in trauma hemorrhage Sirat Khan, MD, Karim Brohi, MD, Manik Chana, MD, Imran Raza, MD, Simon Stanworth, MD, Christine Gaarder, MD, PhD, Ross Davenport, MD, PhD, on behalf of the International Trauma Research Network (INTRN), London, United Kingdom Trauma hemorrhage continues to carry a high mortality rate despite changes in modern practice. Traditional approaches to the massively bleeding patient have been shown to result in persistent coagulopathy, bleeding, and poor outcomes. Hemostatic (or damage control) resuscitation developed from the discovery of acute traumatic coagulopathy and increased recognition of the negative consequences of dilutional coagulopathy. These strategies concentrate on early delivery of coagulation therapy combined with permissive hypotension. The efficacy of hemostatic resuscitation in correcting coagulopathy and restoring tissue perfusion during acute hemorrhage has not been studied. METHODS: This is a prospective cohort study of ROTEM and lactate measurements taken from trauma patients recruited to the multicenter Activation of Coagulation and Inflammation in Trauma (ACIT) study. A blood sample is taken on arrival and during the acute bleeding phase after administration of every 4 U of packed red blood cells (PRBCs), up to 12 U. The quantity of blood products administered within each interval is recorded. RESULTS: Of the 106 study patients receiving at least 4 U of PRBC, 27 received 8 U to 11 U of PRBC and 31 received more than 12 U of PRBC. Average admission lactate was 6.2 mEq/L. Patients with high lactate (Q5 mEq/L) on admission did not clear lactate until hemorrhage control was achieved, and no further PRBC units were required. On admission, 43% of the patients were coagulopathic (clot amplitude at 5 minutes e 35 mm). This increased to 49% by PRBC 4; 62% by PRBC 8 and 68% at PRBC 12. The average fresh frozen plasma/PRBC ratio between intervals was 0.5 for 0 U to 4 U of PRBC, 0.9 for 5 U to 8 U of PRBC, 0.7 for 9 U to 12 U of PRBC. There was no improvement in any ROTEM parameter during ongoing bleeding. CONCLUSION: While hemostatic resuscitation offers several advantages over historical strategies, it still does not achieve correction of hypoperfusion or coagulopathy during the acute phase of trauma hemorrhage. Significant opportunities still exist to improve management and improve outcomes for bleeding trauma patients. (J Trauma Acute Care Surg. 2014;76: 561Y568. Copyright * 2014 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Epidemiologic study, level III. KEY WORDS: Damage control; resuscitation; plasma; transfusion. BACKGROUND:
T
rauma hemorrhage continues to carry a high mortality rate despite changes in modern practice.1,2 Early changes in coagulation, driven by tissue injury and hypoperfusion, are evident on arrival in the emergency department (ED) in 25% of severely injured patients.3 Traditional approaches to the massively bleeding patient have been shown to result in persistent coagulopathy, bleeding, and poor outcomes.4Y6 The concept of hemostatic (or damage-control) resuscitation7 has developed from the discovery of acute traumatic coagulopathy8 and increased recognition of the negative consequences of dilutional coagulopathy.9 These strategies concentrate on the early delivery of coagulation therapy (plasma and platelet transfusions)
Submitted: September 13, 2013, Revised: November 19, 2013, Accepted: November 22, 2013. From the Centre for Trauma Sciences (S.K., K.B., M.C., I.R., R.D.), Blizard Institute, Bart’s & the London School of Medicine, Queen Mary University of London, London; NHS Blood and Transplant (S.S.), John Radcliffe Hospital, Oxford, United Kingdom; and Department of Traumatology (C.G.), Oslo University Hospital Ullevaal, Oslo, Norway. This study was presented at the 72nd annual meeting of the American Association of Surgery for Trauma, September 18Y21, 2013, in San Francisco, California. Address for reprints: Ross Davenport, MD, PhD, Centre for Trauma Sciences, Blizard Institute, Bart’s & the London School of Medicine, Queen Mary University of London, London, United Kingdom; email: [email protected]. DOI: 10.1097/TA.0000000000000146
as part of a massive hemorrhage protocol (MHP) combined with permissive hypotension and early hemostatic procedures. The efficacy of hemostatic resuscitation (HR) to restore tissue perfusion and correct coagulopathy during acute hemorrhage has not been studied. Coagulopathy and hypoperfusion have been shown to correct with HR but only at the end of damage-control surgery or after admission to critical care, presumably after hemorrhage control has been achieved.10 We have studied the bleeding phase and previously shown that fibrinogen, the primary substrate of coagulation, falls during hemorrhage. Despite HR, fibrinogen did not correct with standard damage-control resuscitation.11 While high-dose plasma and platelets have both been associated with improved outcomes,12,13 we reported preliminary findings that different ratios of fresh frozen plasma (FFP) and packed red blood cell (PRBC) had minimal effect on the coagulation profile during trauma hemorrhage.12 It remains unclear whether HR can improve coagulopathy or hypoperfusion during acute bleeding. The overall objective of this study was to determine whether HR achieved its stated aims of correction of coagulopathy and restoration of perfusion during trauma hemorrhage. Our specific aims were to characterize lactate clearance during hemorrhage and to describe the effect of varying FFP/PRBC ratios on
J Trauma Acute Care Surg Volume 76, Number 3
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
561
J Trauma Acute Care Surg Volume 76, Number 3
Khan et al.
correction of trauma induced coagulopathy (TIC) when administered during damage-control resuscitation. We conducted a prospective cohort study of bleeding trauma patients presenting directly to three major international trauma centers.
PATIENTS AND METHODS Study Design Trauma patients presenting directly to three Level I trauma centers between January 2008 and January 2013 were eligible for inclusion. Adult trauma patients who met the local criteria for full trauma team activation were included in the prospective Activation of Coagulation and Inflammation in Trauma (ACIT) observational study. ACIT is an observational study conducted at trauma centers, which are members of the International Trauma Research Network (INTRN). There were three (seven at time of publication) active study sites with data-sharing protocols in place at the time of enrollment for this particular study. Arterial blood was drawn at arrival in the ED and analyzed with laboratory prothrombin time and ROTEM. Patients that continued to exsanguinate had further samples drawn at 4, 8, and 12 PRBC units. Transfusion requirements were recorded during these intervals. Patients requiring four or more PRBC units were analyzed.
Patient Selection All adult trauma patients who met the local criteria for trauma team activation were eligible for enrollment when research personnel were present. Exclusion criteria included arrival in the ED for more than 2 hours following injury; administration of more than 2,000 mL of intravenous fluid before hospital arrival; transfer from another hospital, and burns more than 5% total body surface area. Patients were retrospectively excluded if they did not receive more than 4 U of PRBC, if they declined to give consent for the research study, if they were receiving anticoagulant medications (not including aspirin), or if they had moderate or severe liver disease or a known bleeding diathesis. The majority of included patients were unable to provide informed consent at the time of enrollment, and consent was therefore obtained from the trauma team leader (a physician independent to the research study) who acted as the patient’s legally authorized representative. Written consent from the patient or next of kin was obtained as soon as after the enrollment as appropriate. The study was reviewed and approved by the respective National Research Ethics Committee of each country involved in the study.
Major Hemorrhage Protocol Severely injured trauma patients requiring blood transfusions were routinely managed according to an MHP. The strict activation criteria for the MHP included when a patient presented with a systolic blood pressure of less than 90 mm H, demonstrated a poor response to initial fluid resuscitation, and/or had suspected active hemorrhage. The majority of MHP activations are initiated in the prehospital phase by emergency physicians on scene before the arrival of the patient in the ED. When activated, transfusion begins with PRBCs that are held in a blood fridge within the ED (immediate availability). For London 562
and Oxford, target ratios for FFP/PRBC are 2:3 with platelets and cryoprecipitate provided after every 6 U of PRBC transfused. Delivery is in cooler boxesVPack A contains 4 U of FFP; Pack B and all subsequent packs contain 6 U of PRBC, 2 pools of cryoprecipitate and 1 pool of platelets. In Oslo, FFP/PRBC is provided in a 1:1 ratio (Octaplas) with platelets issued with every 5 U of PRBC and monitored according to conventional coagulation tests.
Sampling Technique A 30-mL research blood sample was drawn from either the femoral artery along with the standard trauma laboratory tests (peripheral blood count, clotting screen, and arterial blood gas) within 20 minutes of arrival in the ED. For those patients with active bleeding who required transfusion, further blood samples were taken after the 4th, 8th, and 12th unit of PRBC were administered and on Days 1 and 3 of admission. Blood for ROTEM analysis was drawn into citrated vacutainers and processed in the trauma research laboratories. Samples for prothrombin ratio (PTr) were collected into citrated vacutainers and processed in the hospital laboratories.
Sample Analysis ROTEM samples were processed within 2 hours of blood draw using a ROTEM delta instrument (TEM International GmbH, Munich, Germany) at 37-C. The methodology and the parameters of ROTEM have been described previously.14 EXTEM assay (tissue factor initiated clotting) was measured at each sampling interval. All pipetting steps and the mixing of reagents with samples were performed with an automated electronic pipette program. Clotting time (CT), clot amplitude at 5 minutes (CA5), > angle, and maximum clot firmness (MCF) were reported for each sample analyzed. Prothrombin time was processed in the hospital laboratories according to standard protocols.
Definition of Coagulopathy and Abnormal Perfusion We defined coagulopathy by ROTEM as a 5-minute EXTEM clot amplitude (CA5) of 35 mm or less.15 This has been shown to accurately identify acute traumatic coagulopathy (ATC) and predicts the need for massive transfusion (defined as Q10 units of PRBC in 24 hours). From this original 300 patient cohort, we classified abnormal ROTEM parameters as T1 SD from normal (PTr G 1.2), CT greater than 94 seconds, CFT greater than 171 seconds, MCF less than 54 mm, > angle less than 65 degrees. A lactate level of 5.0 mmol/L or more was selected to indicate patients with severe hemorrhage.16 We defined normalization of lactate when levels dropped below the upper limit of normal, that is, 2.0 mEq/L.17,18 Completion of PRBC transfusion was used as a surrogate for hemorrhage control.
Data Collection and Analysis Data were collected prospectively on patient demographics, time of injury, mechanism (blunt or penetrating), prehospital fluid administration, time of arrival in the ED, baseline vital signs, and total transfusion requirements in the first 12 hours of admission. For each 4 U of PRBC transfused, interval FFP/PRBC ratios were * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 76, Number 3
Khan et al.
calculated up to the 12th PRBC unit. The change in coagulation parameters within each 4 U of PRBC interval was calculated, for example, Time Point A (baseline) to Time Point B (after the 4th, 8th, and 12th unit of PRBC). Statistical analysis was performed using GraphPad PRISM version 5 (GraphPad Software Inc., San Diego CA) and Microsoft Excel 2007 (Microsoft, Inc., Redmond, WA). Normal quantile plots were used to test for normal distribution. Parametric data are expressed as mean T 95% confidence intervals. Nonparametric data are given as median (interquartile range [IQR]). A p G 0I05 was chosen to represent statistical significance throughout.
RESULTS A total of 810 patients were included in the ACIT study during the 52-month period with 49 exclusionsVconsent declined or not possible (21 patients) and retrospective exclusion criteria (28 patients). A total of 106 study patients were transfused four or more units of PRBC during the study period, with 27 receiving 8 U to 11 U of PRBC and 31 who received 12 or more PRBC units. The average FFP/PRBC ratio between intervals was 0.5 for 0 U to 4 U of PRBC, 0.9 for 5 U to 8 U of PRBC, 0.7 for 9 U to 12 U of PRBC, and 0.7 at Hour 12. Median crystalloid (CSL) use was 0 mL for all intervals: 0 U to 4 U of PRBC (IQR, 0Y100 mL), 5 U to 8 U of PRBC (IQR, 0Y500 mL), 9 U to 12 U of PRBC (IQR, 0Y1,000 mL). Overall mortality was 35%. Early deaths during the bleeding intervals were 10 (10%) for 5 U to 8 U of PRBC, 6 (6%) for 9 U to 12 U of PRBC, 6 (6%) for 12 U or more PRBCs to Hour 12. Clinical characteristics, admission physiology, and laboratory parameters are detailed in Table 1. Average admission lactate was 6.2 mEq/L. Increased transfusion requirement were associated with higher baseline lactates (4Y7 U of PRBC, lactate, 3.2 mEq/L vs. 912 U of PRBC, lactate 9.2 mEq/L, p G 0.05). Lactate levels remained elevated during the bleeding episode regardless of overall transfusion (Fig. 1AYC). Patients who received up to 7 U of PRBC did not correct lactate levels until transfusion was complete (Fig. 1A). Similarly, patients requiring 8 U to 11 U of PRBC (Fig. 1B) or 12 U or more PRBCs (Fig. 1C) did not clear lactate during hemorrhage. Lactate levels were only normal (and significantly lower) on the Day 1 sample (Fig. 1AYC). On admission, 43% of patients were coagulopathic (CA5 e35 mm). More patients became coagulopathic as hemorrhage continued and increased to 49% by 4 U of PRBC, 62% by 8 U of PRBC, and 68% at 12 U of PRBC. There was no improvement in any ROTEM parameter during ongoing bleeding. Average values remained below the diagnostic threshold for coagulopathy (CA5 e 35 mm) throughout the bleeding episode (Fig. 2AYC) with only partial correction evident at Day 1. During trauma hemorrhage, there was profound variability in individual response to HR (Fig. 2D). Deterioration in MCF was also observed during ongoing bleeding (Fig. 3AYC). In the highest transfusion group, MCF deteriorated 15% during HR and only improved on Day 1 (Q12 U of PRBC: Time 0 MCF, 46 mm vs. Day 1 MCF, 59 mm; p G 0.05) (Fig. 3C). Again, there was significant variability to HR (Fig. 3D).
TABLE 1. Clinical Characteristics of Patients Patients
n = 106
Age, y Male Injury Injury Severity Score (ISS) ISS 9 15 Blunt injury Admission parameters Systolic blood pressure G 90 mm Hg Base deficit 9 6 mEq/L Lactate 9 5 mEq/L PT, s PTr 9 1.2 CA5 e 35, mm Damage-control procedure Operating room Interventional radiology Combined operating room/interventional radiology Fluid requirements (first 12 h) Prehospital crystalloid, mL Crystalloid, mL Colloid, mL PRBC units PRBC Q 10 U FFP units Platelet units Cryoprecipitate units
44 (30Y60) 75 (76) 34 (25Y41) 99 (94) 92 (87) 41 (46) 68 (67) 44 (42) 12.9 (11.7Y14.6) 40 (40) 41 (43) 61 (70) 6 (7) 4 (5) 500 (100Y3,000) 2,500 (1,300Y4,000) 875 (0Y1,688) 8 (5Y13) 40 (38) 6 (4Y8) 1 (1Y2) 2 (0Y2)
Values are given as n (%) or median (IQR).
Prolonged CT (994 seconds) was observed in 33 patients (31%); despite HR, this parameter also did not correct during ongoing bleeding (Fig. 4AYD), with correction only observed at Day 1. A total of 65 patients (61%) had an abnormally flattened > angle of less than 64 degrees (normal range, 63Y81 degrees) and similar to the clot strength parameters deteriorated throughout the bleeding episode (Time 0, > = 59 degrees; 4Y12 U of PRBC, > = 54Y55 degrees; Day 1, > = 70 degrees (p G 0.05).
DISCUSSION In this multicenter, prospective study of severely injured, bleeding patients, we have shown that HR does not correct hypoperfusion or coagulopathy during the acute phase of trauma hemorrhage. Transfusion of FFP/PRBC ratios greater than 1:2 throughout the bleeding episode failed to normalize any ROTEM parameters and lactic acidemia was not cleared until Day 1, after hemorrhage control had been achieved. Serial serum lactate measurements are a prognostic marker of tissue hypoperfusion and severity of hemorrhagic shock in trauma,16,19 and clearance of lactate is correlated with survival and organ failure.20 Correction of lactic acidemia remains an important therapeutic goal of HR and damage-control surgery.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
563
Khan et al.
J Trauma Acute Care Surg Volume 76, Number 3
Figure 1. Lactate clearance during hemorrhage. AYC box and whisker plots, median, IQR and adjusted range. *Versus Time 0. A, Patients receiving 4 U to 7 U of PRBC. B, Patients receiving 8 U to 11 U of PRBC. C, Patients receiving 12 U or more PRBCs.
Figure 2. Change in CA5 (mm) during bleeding episode in coagulopathic patients (CA5 e 35 mm) stratified by transfusion requirements. Dotted line is diagnostic threshold for ATC (CA5 e 35 mm). A, Patients receiving 4 U to 7 U of PRBC. Time 0: CA5, 35 mm versus Day 1: CA5, 39 mm (p G 0.05). B, Patients receiving 8 U to 11 U of PRBC. Time 0: CA5, 33 mm versus Day 1: CA5, 32 mm (p = 0.78). C, Patients receiving 12 U or more PRBCs. Time 0: CA5, 27 mm versus Day 1: CA5, 39 mm (p G 0.05). D, Individual response in all patients receiving four or more PRBC units. 564
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 76, Number 3
Khan et al.
Figure 3. Change in MCF (mm) during bleeding episode in coagulopathic patients (MCF e 54 mm) stratified by transfusion requirements. Dotted line is the diagnostic threshold for ATC (MCF e 54 mm). A, Patients receiving 4 U to 7 U of PRBC. Time 0: MCF, 55 mm versus Day 1: MCF, 58 mm (nonsignificant). B, Patients receiving 8 U to 11 U of PRBC. Time 0: MCF, 51 mm versus Day 1: MCF, 53 mm (nonsignificant). C, Patients receiving 12 U or more PRBCs. Time 0: MCF, 46 mm versus Day 1: MCF, 59 mm (p G 0.05). D, Individual response in all patients receiving four or more PRBC units.
Figure 4. Change in CT (seconds) during bleeding episode in coagulopathic patients (CT 9 94 seconds) stratified by transfusion requirements. Dotted line is diagnostic threshold for ATC (CT 9 94 seconds). A, Patients receiving 4 U to 7 U of PRBC. Time 0: CT, 108 seconds versus Day 1: CT, 69 seconds (p G 0.05). B, Patients receiving 8 U to 11 U of PRBC. Time 0: CT, 107 seconds versus Day 1: CT, 69 seconds (p = 0.07). C, Patients receiving 12 U or more PRBCs. Time 0: CT, 126 seconds versus Day 1: CT, 63 seconds (p G 0.05). D, Individual response in all patients receiving four or more PRBC units. * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
565
J Trauma Acute Care Surg Volume 76, Number 3
Khan et al.
Combination of these strategies in combat casualties has been shown to correct metabolic derangements in a before and after study of hemorrhage control.10 Our study is consistent with this, but we have shown that restoration of normal tissue perfusion and improvement in TIC does not occur until hemorrhage control is achieved. Clot strength and CTs deteriorated during ongoing bleeding and only corrected when no further PRBC units were required. Our findings reinforce those from a comparative study of laboratory-guided MHP versus empiric 1:1:1 (PRBC/FFP/ platelets) protocol.13 Fibrinogen, platelets, and prothrombin time trends in the acute bleeding episode were identical in both transfusion strategies, despite a fourfold increase in patients receiving high FFP/PRBC ratio (91:2). HR may offer several advantages over historical transfusion strategies, for example, rapid access to empiric component therapy, but the findings from this study show it has limited efficacy in correcting TIC. The benefits of HR have been postulated to a reduction in CSL use. Studies have shown that increased CSL use is associated with dilutional coagulopathy and worse outcomes.9 Permissive hypotension was practiced at all study sites with limited volumes of CSL infused before baseline blood samples. More severe changes in coagulopathy may have been seen had HR protocols not been followed by centers in this study. HR may therefore protect somewhat from deterioration of these parameters during bleeding but in itself is not able to correct TIC. There are several limitations to this study. We were only able to sample up to the 12th PRBC unit transfused and therefore could not follow coagulation profile trends up to the point of hemorrhage control in some patients. Precise timing of hemorrhage control is difficult to ascertain, and therefore, we used completion of PRBC transfusion as a surrogate measure for control of bleeding. Average FFP/PRBC ratios were greater than 1:2 throughout, and we did not look specifically at the effect of differential ratios on coagulation or hypoperfusion. We also did not look at other component therapies (platelets and cryoprecipitate), although clearly, they will influence correction of TIC. In particular, fibrinogen is fundamental to hemostasis and falls to critical levels soon after the onset of major trauma hemorrhage.11 Early aggressive fibrinogen replacement may therefore prove efficacious in the correction of TIC and merits further investigation. We did not examine the fibrinolytic component of TIC in this study, although from 2010 and the publication of the CRASH-2 trial, MHPs at all sites specified early administration of tranexamic acid. Undertreated or persistent hyperfibrinolysis may contribute to poor clot function and is therefore part of continued investigation within INTRN. This investigation questions the therapeutic mechanism and efficacy of HR in the acute phase of trauma hemorrhage. Aggressive high-volume replacement of plasma has little effect on deranged ROTEM coagulation parameters and hypoperfusion. Arrest of hemorrhage and cessation of transfusion seem necessary before TIC can be corrected and tissue perfusion restored. Following hemorrhage control and completion of transfusion, ROTEM parameters do not immediately return to values seen in normal healthy controls. The question for future design of goal-directed therapy, using devices such as ROTEM remains the identification of target values for coagulation 566
parameters in trauma hemorrhage. Significant opportunities exist to tailor management and improve outcomes for bleeding trauma patients. A balanced and potentially novel strategy is required to achieve correction of TIC while maintaining organ perfusion. HR as practiced by the institutions in this study may be protective but is neither hemostatic nor resuscitative. AUTHORSHIP S.K., K.B., and R.D. conceived the study. R.D., S.K., I.R., and M.C. collected and analyzed the data. R.D., C.G., S.S., and K.B. wrote and edited the manuscript.
DISCLOSURE TEM Innovations (ROTEM) provided unrestricted support in the form of equipment and reagent support for ACIT study. S.K. and R.D. have received honorarium as invited speakers. This study was funded in part by the National Institute for Health Research (UK) Program Grant for Applied Research (RP-PG-0407-10036).
REFERENCES 1. Holcomb JB, Del Junco DJ, Fox EE, Wade CE, Cohen MJ, Schreiber MA, Alarcon LH, Bai Y, Brasel KJ, Bulger EM, et al. The Prospective, Observational, Multicenter, Major Trauma Transfusion (PROMMTT) study: comparative effectiveness of a time-varying treatment with competing risks. JAMA Surg. 2013;148(2):127Y136. 2. Mitra B, Cameron PA, Mori A, Fitzgerald M. Acute coagulopathy and early deaths post major trauma. Injury. 2012;43(1):22Y25. 3. Brohi K, Cohen MJ, Davenport RA. Acute coagulopathy of trauma: mechanism, identification and effect. Curr Opin Crit Care. 2007;13(6):680Y685. 4. Gonzalez EA, Moore FA, Holcomb JB, Miller CC, Kozar RA, Todd SR, Cocanour CS, Balldin BC, McKinley BA. Fresh frozen plasma should be given earlier to patients requiring massive transfusion. J Trauma. 2007; 62(1):112Y119. 5. Murad MH, Stubbs JR, Gandhi MJ, Wang AT, Paul A, Erwin PJ, Montori VM, Roback JD. The effect of plasma transfusion on morbidity and mortality: a systematic review and meta-analysis. Transfusion. 2010; 50(6):1370Y1383. 6. Ho AM, Karmakar MK, Dion PW. Are we giving enough coagulation factors during major trauma resuscitation? Am J Surg. 2005;190(3):479Y484. 7. Holcomb JB, Jenkins D, Rhee P, Johannigman J, Mahoney P, Mehta S, Cox ED, Gehrke MJ, Beilman GJ, Schreiber M, et al. Damage control resuscitation: directly addressing the early coagulopathy of trauma. J Trauma. 2007;62(2):307Y310. 8. Brohi K, Singh J, Heron M, Coats T. Acute traumatic coagulopathy. J Trauma. 2003;54(6):1127Y1130. 9. Maegele M, Lefering R, Yucel N, Tjardes T, Rixen D, Paffrath T, et al. Early coagulopathy in multiple injury: an analysis from the German Trauma Registry on 8724 patients. Injury. 2007;38(3):298Y304. 10. Morrison JJ, Ross JD, Poon H, Midwinter MJ, Jansen JO. Intra-operative correction of acidosis, coagulopathy and hypothermia in combat casualties with severe haemorrhagic shock. Anaesthesia. 2013;68(8):846Y850. 11. Rourke C, Curry N, Khan S, Taylor R, Raza I, Davenport R, Stanworth S, Brohi K. Fibrinogen levels during trauma hemorrhage, response to replacement therapy, and association with patient outcomes. J Thromb Haemost. 2012;10(7):1342Y1351. 12. Davenport R, Curry N, Manson J, De’Ath H, Coates A, Rourke C, Pearse R, Stanworth S, Brohi K. Hemostatic effects of fresh frozen plasma may be maximal at red cell ratios of 1:2. J Trauma. 2011;70(1):90Y95; discussion 95Y96. 13. Chambers LA, Chow SJ, Shaffer LE. Frequency and characteristics of coagulopathy in trauma patients treated with a low- or high-plasma-content massive transfusion protocol. Am J Clin Pathol. 2011;136(3):364Y370.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 76, Number 3
Khan et al.
14. Ganter MT, Hofer CK. Coagulation monitoring: current techniques and clinical use of viscoelastic point-of-care coagulation devices. Anesth Analg. 2008;106(5):1366Y1375. 15. Davenport R, Manson J, De’Ath H, Platton S, Coates A, Allard S, Hart D, Pearse R, Pasi KJ, MacCallum P, et al. Functional definition and characterization of acute traumatic coagulopathy. Crit Care Med. 2011;39(12): 2652Y2658. 16. Re´gnier MA, Raux M, Le Manach Y, Asencio Y, Gaillard J, Devilliers C, Langeron O, Riou B. Prognostic significance of blood lactate and lactate clearance in trauma patients. Anesthesiology. 2012;117(6): 1276Y1288. 17. Husain FA, Martin MJ, Mullenix PS, Steele SR, Elliott DC. Serum lactate and base deficit as predictors of mortality and morbidity. Am J Surg. 2003;185(5):485Y491. 18. Abramson D, Scalea TM, Hitchcock R, Trooskin SZ, Henry SM, Greenspan J. Lactate clearance and survival following injury. J Trauma. 1993;35(4):584Y588; discussion 588Y589. 19. Odom SR, Howell MD, Silva GS, Nielsen VM, Gupta A, Shapiro NI, Talmor D. Lactate clearance as a predictor of mortality in trauma patients. J Trauma Acute Care Surg. 2013;74(4):999Y1004. 20. Manikis P, Jankowski S, Zhang H, Kahn RJ, Vincent JL. Correlation of serial blood lactate levels to organ failure and mortality after trauma. Am J Emerg Med. 1995;13(6):619Y622.
DISCUSSION Dr. Yashuhiro Otomo (Tokyo, Japan): The combination of fibrinolysis dilution and inadequate function produces a complex coagulopathy in many trauma patients who receive massive transfusions. An aggressive transfusion therapy with plasma and platelet could prevent or correct deficiencies and, thereby, might improve survival. Indeed, multiple trauma programs have achieved dramatically increased survival rate. However, most of these studies are retrospective, case controlled review of notoriously complex patient situations with many injuries and treatment features that affect outcome. Dr. Davenport and coauthors have conducted this prospective cohort study of bleeding trauma patients presenting directly to three major international trauma centers. This study was well designed and organized and successfully collected a large amount data at each time point. The authors may predict a positive effect of aggressive transfusion of plasma on the outcome and also aimed to describe the effect of varying FFP:RBC ratios on correction of coagulopathy. However, surprisingly, in this multi-center prospective study hemostatic resuscitation was not effective to correct hypoperfusion or coagulopathy during the acute phase of trauma hemorrhage. From these results the authors made this provocative title: Hemostatic Resuscitation is Neither Hemostatic Nor Resuscitative in Trauma Hemorrhage. This is the first prospective study to show an evidence that aggressive transfusion of plasma have very few effects in reversing coagulopathy caused by severe trauma bleeding. I want to commend the excellent study group for this great achievement. I have several comments and questions for the authors. First, the definition of hemostatic resuscitation is not clear. My understanding of hemostatic resuscitation is the combination of avoiding dilution and hypothermia, enough amount of administration of FFP, administration of fibrinogen, PCC or cryoprecipitate and administration of anti-hypofibrinolytic agent.
In your study you should clearly describe the definition of hemostatic resuscitation. Second, among the total 106 patients that you included into your analysis you mentioned overall 69that is, 55% of patientswere coagulopathic by reduced clot strength, CA5 less than 36 millimeters. This means 35% of patients included in your analysis are not coagulopathic. I believe that this study should include only coagulopathic patients or you should show the data analyzed only on the 35%non-coagulopathic patients. Third, in your study, the timing that hemorrhage control was achieved is not clearly described. From your presentation that timing appears to be the Day 1; however, I assume that that timing would be different from case to case, such as within two hours of the injury or 24 hours after injury. Please make it clear. Fourth, you didn’t mention about the survival bias. As you know, the clinical researches on transfusion ratios, one of the most important issue that should be addressed has been this survival bias. In order to evaluate the efficacy of aggressive FFP transfusion you should touch upon it. Finally, a massive transfusion of FFP provides a massive amount of protein C to the patient. According to your theory this would cause a large amount of production of activated protein C that will result in a progression of trauma-induced coagulopathy. Why did you predict that high ratio of FFP transfusion might be effective for trauma massive hemorrhage? That is competing against your theory of trauma-induced coagulopathy. The authors said a balanced and potentially novel strategy is required. I completely agree with that. I would like to thank the association for the privilege of the floor. Dr. John B. Holcomb (Houston, Texas): Dr. Davenport, I congratulate you on a very clever title. Dr. Brohi’s favorite term, I think. The problem with this paper is it is observational. And so deriving over reaching conclusions with no control group greatly overstates your data. Importantly, you nicely presented a lot of laboratory data but you didn’t tell us how the patients did. Probably the most important missing element here is time. All of this interventions are extraordinarily timedependent. When did the patients die? What did they die from. How did they do? I congratulate you for partially adopting the hemostatic or damage control methods by limiting crystalloid and then trying to achieve an early ratio. But in fact you didn’t. You didn’t start plasma early. You caught up late. The PROMMTT trial of 1200 patients recently published several papers in the Journal of Trauma. One of them directly addressed this issue: lack of plasma early, catching up later, resulted in poor outcomes. Please comment. Dr. Richard J. Mullins (Portland, Oregon): This study raises the question: is there surgeon variation that accounts for your outcome? Your study shows that massive transfusion has associated with a variation in death but how have you accounted for the fact that the patients who had four units of blood were operated on by a master surgeon like Rao Ivatury and the patients who got 12 units of blood were operated on by somebody who occasionally takes trauma call?
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
567
J Trauma Acute Care Surg Volume 76, Number 3
Khan et al.
The general principle is when you are going to talk about bleeding to death I don’t know how you are going to account for the fact that there is in fact variation in the ability of surgeons to stop bleeding that has a major bias in your data. Dr. Eileen M. Bulger (Seattle, Washington): I think the fundamental question you are asking is unfair. I mean I think there is no question that we have to control bleeding to correct coagulopathy and if you can’t control bleeding your patient is going to die. So as much resuscitation as we do, we’re trying to keep up and not to completely correct coagulopathy during the time of bleeding. I think that’s an unfair goal. I also want to echo Dr. Holcomb’s concerns about timing and survival bias because absolutely dividing patients by the amount of blood they got, if they die early and they don’t get a lot of blood, it’s a very mixed group in that low blood use group so you have to account for that in the analysis. And I really want to know how you accounted for early deaths in this study and how you accounted for these timing issues of people who achieve hemostasis at different times and people who get ratios at different times. Dr. Mitchell Jay Cohen (San Francisco, California): I echo the comments that have been previously stated. I think this paper begins to address one of the central issues in hemostatic resuscitation which is whether these patients do better from hemostatic resuscitation because they stop bleeding and don’t bleed to death or do they do better because there is something else in the ‘‘special sauce’’ that is keeping the patients and, their endothelium happy and keeping them alive? So my question to you is what else is in your data? First of all, did these patients actually have better outcomes? What does the outcome data look like when they got hemostatic resuscitation? And if it’s not bleeding and it’s not lactate clearance, what else is going on, in your opinion? Dr. Zsolt Balogh (Newcastle, Australia): Congratulations, especially supporting my bias on eliminating crystalloids is probably the cause for better outcomes. I must highlight that damage control resuscitation was introduced based on much lower quality studies than this one, so I don’t think it is too bad. My question is how long was the time frame of hypotension of these patients and what time you managed to catch up with the desired goal of blood and FFP during the resuscitation? Dr. Louis Alarcon (Pittsburgh, Pennsylvania): Two questions, please. First, did you include tranexamic acid in your massive transfusion protocol? And the second question relates to the PROMMTT study, wherein we learned that the ratios of blood products varied over time in any particular patient. So my questions are: at what time point did you calculate the plasma and platelet to RBC ratios? And how did you account for the fact that these ratios, in any particular patient, vary considerably over time? Thank you. Dr. Ross Davenport (London, United Kingdom): Thank you all for those insightful comments. Clearly the title was
568
provocative and is designed to generate discussion and to consider alternative hypotheses. Firstly, to answer some of Professor Otomo’s comments. All the centers adhered to the principles of hemostatic resuscitation. Fibrinogen supplementation was administered in the form of cryoprecipitate in addition to platelets. All centers now use tranexamic acid. Some of the patients were involved in this study prior to the publication of the CRASH-2 trial, therefore, not all of the patients received TXA although the majority did. We only included coagulopathic patients because we felt that there was limited benefit in examining those who are not coagulopathic as FFP would not be expected to alter their coagulation profiles. The timing of hemorrhage control is an important and vexing question and the comments made by Dr. Holcomb are very important. Clearly, the more shocked a patient is or the longer they remain shocked the worse the coagulopathy is likely to be and further drive the bleeding, shock and coagulopathic process. All patients in this study had damage control resuscitation by attendings who only take trauma calls. It is difficult to ascertain the precise time to hemorrhage control. However all these patients had a massive hemorrhage protocol would have had control of hemorrhage within the first one-to-two hours of arrival in the emergency department. We didn’t look specifically at the effects of different ratios in this study. We have previously shown that differential ratios of FFP to packed red cells have minimum effect on the coagulation profile. The maximum benefit is seen at a ratio of one to two. Clearly, this is something that does require further examination. And I think we have to think of newer ways to actually work out what rapid bleeding means and how to precisely evaluate the effects of other component therapies, clearly, platelets and fibrinogen supplementation. You asked for my comments on tranexamic acid. It’s always a topic of controversy when we come to these sorts of meetings. As I’ve said, the patients included in this study were both before and after 2010 (CRASH-2). There are clearly many questions that still need to be answered regarding tranexamic acid, certainly how it actually works, the dosing, optimal timing, and its precise role in a mature trauma system. Lastly, the comments about activated protein C. We have shown that this is a potential mediator of this coagulopathy. Those patients that received high doses of plasma could theoretically, further drive generation of activated protein C and coagulopathy. Regarding limitation of crystalloids and actually what FFP does, I think this is an important question. There are numerous factors in FFP which we don’t know, what are their effects on coagulation or inflammation and whether or not it acts early or late, for example in stabilization of the endothelial glycocalyx. Thank you.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Hemostatic resuscitation is neither hemostatic nor resuscitative in trauma hemorrhage Sirat Khan, MD, Karim Brohi, MD, Manik Chana, MD, Imran Raza, MD, Simon Stanworth, MD, Christine Gaarder, MD, PhD, Ross Davenport, MD, PhD, on behalf of the International Trauma Research Network (INTRN), London, United Kingdom Trauma hemorrhage continues to carry a high mortality rate despite changes in modern practice. Traditional approaches to the massively bleeding patient have been shown to result in persistent coagulopathy, bleeding, and poor outcomes. Hemostatic (or damage control) resuscitation developed from the discovery of acute traumatic coagulopathy and increased recognition of the negative consequences of dilutional coagulopathy. These strategies concentrate on early delivery of coagulation therapy combined with permissive hypotension. The efficacy of hemostatic resuscitation in correcting coagulopathy and restoring tissue perfusion during acute hemorrhage has not been studied. METHODS: This is a prospective cohort study of ROTEM and lactate measurements taken from trauma patients recruited to the multicenter Activation of Coagulation and Inflammation in Trauma (ACIT) study. A blood sample is taken on arrival and during the acute bleeding phase after administration of every 4 U of packed red blood cells (PRBCs), up to 12 U. The quantity of blood products administered within each interval is recorded. RESULTS: Of the 106 study patients receiving at least 4 U of PRBC, 27 received 8 U to 11 U of PRBC and 31 received more than 12 U of PRBC. Average admission lactate was 6.2 mEq/L. Patients with high lactate (Q5 mEq/L) on admission did not clear lactate until hemorrhage control was achieved, and no further PRBC units were required. On admission, 43% of the patients were coagulopathic (clot amplitude at 5 minutes e 35 mm). This increased to 49% by PRBC 4; 62% by PRBC 8 and 68% at PRBC 12. The average fresh frozen plasma/PRBC ratio between intervals was 0.5 for 0 U to 4 U of PRBC, 0.9 for 5 U to 8 U of PRBC, 0.7 for 9 U to 12 U of PRBC. There was no improvement in any ROTEM parameter during ongoing bleeding. CONCLUSION: While hemostatic resuscitation offers several advantages over historical strategies, it still does not achieve correction of hypoperfusion or coagulopathy during the acute phase of trauma hemorrhage. Significant opportunities still exist to improve management and improve outcomes for bleeding trauma patients. (J Trauma Acute Care Surg. 2014;76: 561Y568. Copyright * 2014 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Epidemiologic study, level III. KEY WORDS: Damage control; resuscitation; plasma; transfusion. BACKGROUND:
T
rauma hemorrhage continues to carry a high mortality rate despite changes in modern practice.1,2 Early changes in coagulation, driven by tissue injury and hypoperfusion, are evident on arrival in the emergency department (ED) in 25% of severely injured patients.3 Traditional approaches to the massively bleeding patient have been shown to result in persistent coagulopathy, bleeding, and poor outcomes.4Y6 The concept of hemostatic (or damage-control) resuscitation7 has developed from the discovery of acute traumatic coagulopathy8 and increased recognition of the negative consequences of dilutional coagulopathy.9 These strategies concentrate on the early delivery of coagulation therapy (plasma and platelet transfusions)
Submitted: September 13, 2013, Revised: November 19, 2013, Accepted: November 22, 2013. From the Centre for Trauma Sciences (S.K., K.B., M.C., I.R., R.D.), Blizard Institute, Bart’s & the London School of Medicine, Queen Mary University of London, London; NHS Blood and Transplant (S.S.), John Radcliffe Hospital, Oxford, United Kingdom; and Department of Traumatology (C.G.), Oslo University Hospital Ullevaal, Oslo, Norway. This study was presented at the 72nd annual meeting of the American Association of Surgery for Trauma, September 18Y21, 2013, in San Francisco, California. Address for reprints: Ross Davenport, MD, PhD, Centre for Trauma Sciences, Blizard Institute, Bart’s & the London School of Medicine, Queen Mary University of London, London, United Kingdom; email: [email protected]. DOI: 10.1097/TA.0000000000000146
as part of a massive hemorrhage protocol (MHP) combined with permissive hypotension and early hemostatic procedures. The efficacy of hemostatic resuscitation (HR) to restore tissue perfusion and correct coagulopathy during acute hemorrhage has not been studied. Coagulopathy and hypoperfusion have been shown to correct with HR but only at the end of damage-control surgery or after admission to critical care, presumably after hemorrhage control has been achieved.10 We have studied the bleeding phase and previously shown that fibrinogen, the primary substrate of coagulation, falls during hemorrhage. Despite HR, fibrinogen did not correct with standard damage-control resuscitation.11 While high-dose plasma and platelets have both been associated with improved outcomes,12,13 we reported preliminary findings that different ratios of fresh frozen plasma (FFP) and packed red blood cell (PRBC) had minimal effect on the coagulation profile during trauma hemorrhage.12 It remains unclear whether HR can improve coagulopathy or hypoperfusion during acute bleeding. The overall objective of this study was to determine whether HR achieved its stated aims of correction of coagulopathy and restoration of perfusion during trauma hemorrhage. Our specific aims were to characterize lactate clearance during hemorrhage and to describe the effect of varying FFP/PRBC ratios on
J Trauma Acute Care Surg Volume 76, Number 3
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
561
J Trauma Acute Care Surg Volume 76, Number 3
Khan et al.
correction of trauma induced coagulopathy (TIC) when administered during damage-control resuscitation. We conducted a prospective cohort study of bleeding trauma patients presenting directly to three major international trauma centers.
PATIENTS AND METHODS Study Design Trauma patients presenting directly to three Level I trauma centers between January 2008 and January 2013 were eligible for inclusion. Adult trauma patients who met the local criteria for full trauma team activation were included in the prospective Activation of Coagulation and Inflammation in Trauma (ACIT) observational study. ACIT is an observational study conducted at trauma centers, which are members of the International Trauma Research Network (INTRN). There were three (seven at time of publication) active study sites with data-sharing protocols in place at the time of enrollment for this particular study. Arterial blood was drawn at arrival in the ED and analyzed with laboratory prothrombin time and ROTEM. Patients that continued to exsanguinate had further samples drawn at 4, 8, and 12 PRBC units. Transfusion requirements were recorded during these intervals. Patients requiring four or more PRBC units were analyzed.
Patient Selection All adult trauma patients who met the local criteria for trauma team activation were eligible for enrollment when research personnel were present. Exclusion criteria included arrival in the ED for more than 2 hours following injury; administration of more than 2,000 mL of intravenous fluid before hospital arrival; transfer from another hospital, and burns more than 5% total body surface area. Patients were retrospectively excluded if they did not receive more than 4 U of PRBC, if they declined to give consent for the research study, if they were receiving anticoagulant medications (not including aspirin), or if they had moderate or severe liver disease or a known bleeding diathesis. The majority of included patients were unable to provide informed consent at the time of enrollment, and consent was therefore obtained from the trauma team leader (a physician independent to the research study) who acted as the patient’s legally authorized representative. Written consent from the patient or next of kin was obtained as soon as after the enrollment as appropriate. The study was reviewed and approved by the respective National Research Ethics Committee of each country involved in the study.
Major Hemorrhage Protocol Severely injured trauma patients requiring blood transfusions were routinely managed according to an MHP. The strict activation criteria for the MHP included when a patient presented with a systolic blood pressure of less than 90 mm H, demonstrated a poor response to initial fluid resuscitation, and/or had suspected active hemorrhage. The majority of MHP activations are initiated in the prehospital phase by emergency physicians on scene before the arrival of the patient in the ED. When activated, transfusion begins with PRBCs that are held in a blood fridge within the ED (immediate availability). For London 562
and Oxford, target ratios for FFP/PRBC are 2:3 with platelets and cryoprecipitate provided after every 6 U of PRBC transfused. Delivery is in cooler boxesVPack A contains 4 U of FFP; Pack B and all subsequent packs contain 6 U of PRBC, 2 pools of cryoprecipitate and 1 pool of platelets. In Oslo, FFP/PRBC is provided in a 1:1 ratio (Octaplas) with platelets issued with every 5 U of PRBC and monitored according to conventional coagulation tests.
Sampling Technique A 30-mL research blood sample was drawn from either the femoral artery along with the standard trauma laboratory tests (peripheral blood count, clotting screen, and arterial blood gas) within 20 minutes of arrival in the ED. For those patients with active bleeding who required transfusion, further blood samples were taken after the 4th, 8th, and 12th unit of PRBC were administered and on Days 1 and 3 of admission. Blood for ROTEM analysis was drawn into citrated vacutainers and processed in the trauma research laboratories. Samples for prothrombin ratio (PTr) were collected into citrated vacutainers and processed in the hospital laboratories.
Sample Analysis ROTEM samples were processed within 2 hours of blood draw using a ROTEM delta instrument (TEM International GmbH, Munich, Germany) at 37-C. The methodology and the parameters of ROTEM have been described previously.14 EXTEM assay (tissue factor initiated clotting) was measured at each sampling interval. All pipetting steps and the mixing of reagents with samples were performed with an automated electronic pipette program. Clotting time (CT), clot amplitude at 5 minutes (CA5), > angle, and maximum clot firmness (MCF) were reported for each sample analyzed. Prothrombin time was processed in the hospital laboratories according to standard protocols.
Definition of Coagulopathy and Abnormal Perfusion We defined coagulopathy by ROTEM as a 5-minute EXTEM clot amplitude (CA5) of 35 mm or less.15 This has been shown to accurately identify acute traumatic coagulopathy (ATC) and predicts the need for massive transfusion (defined as Q10 units of PRBC in 24 hours). From this original 300 patient cohort, we classified abnormal ROTEM parameters as T1 SD from normal (PTr G 1.2), CT greater than 94 seconds, CFT greater than 171 seconds, MCF less than 54 mm, > angle less than 65 degrees. A lactate level of 5.0 mmol/L or more was selected to indicate patients with severe hemorrhage.16 We defined normalization of lactate when levels dropped below the upper limit of normal, that is, 2.0 mEq/L.17,18 Completion of PRBC transfusion was used as a surrogate for hemorrhage control.
Data Collection and Analysis Data were collected prospectively on patient demographics, time of injury, mechanism (blunt or penetrating), prehospital fluid administration, time of arrival in the ED, baseline vital signs, and total transfusion requirements in the first 12 hours of admission. For each 4 U of PRBC transfused, interval FFP/PRBC ratios were * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 76, Number 3
Khan et al.
calculated up to the 12th PRBC unit. The change in coagulation parameters within each 4 U of PRBC interval was calculated, for example, Time Point A (baseline) to Time Point B (after the 4th, 8th, and 12th unit of PRBC). Statistical analysis was performed using GraphPad PRISM version 5 (GraphPad Software Inc., San Diego CA) and Microsoft Excel 2007 (Microsoft, Inc., Redmond, WA). Normal quantile plots were used to test for normal distribution. Parametric data are expressed as mean T 95% confidence intervals. Nonparametric data are given as median (interquartile range [IQR]). A p G 0I05 was chosen to represent statistical significance throughout.
RESULTS A total of 810 patients were included in the ACIT study during the 52-month period with 49 exclusionsVconsent declined or not possible (21 patients) and retrospective exclusion criteria (28 patients). A total of 106 study patients were transfused four or more units of PRBC during the study period, with 27 receiving 8 U to 11 U of PRBC and 31 who received 12 or more PRBC units. The average FFP/PRBC ratio between intervals was 0.5 for 0 U to 4 U of PRBC, 0.9 for 5 U to 8 U of PRBC, 0.7 for 9 U to 12 U of PRBC, and 0.7 at Hour 12. Median crystalloid (CSL) use was 0 mL for all intervals: 0 U to 4 U of PRBC (IQR, 0Y100 mL), 5 U to 8 U of PRBC (IQR, 0Y500 mL), 9 U to 12 U of PRBC (IQR, 0Y1,000 mL). Overall mortality was 35%. Early deaths during the bleeding intervals were 10 (10%) for 5 U to 8 U of PRBC, 6 (6%) for 9 U to 12 U of PRBC, 6 (6%) for 12 U or more PRBCs to Hour 12. Clinical characteristics, admission physiology, and laboratory parameters are detailed in Table 1. Average admission lactate was 6.2 mEq/L. Increased transfusion requirement were associated with higher baseline lactates (4Y7 U of PRBC, lactate, 3.2 mEq/L vs. 912 U of PRBC, lactate 9.2 mEq/L, p G 0.05). Lactate levels remained elevated during the bleeding episode regardless of overall transfusion (Fig. 1AYC). Patients who received up to 7 U of PRBC did not correct lactate levels until transfusion was complete (Fig. 1A). Similarly, patients requiring 8 U to 11 U of PRBC (Fig. 1B) or 12 U or more PRBCs (Fig. 1C) did not clear lactate during hemorrhage. Lactate levels were only normal (and significantly lower) on the Day 1 sample (Fig. 1AYC). On admission, 43% of patients were coagulopathic (CA5 e35 mm). More patients became coagulopathic as hemorrhage continued and increased to 49% by 4 U of PRBC, 62% by 8 U of PRBC, and 68% at 12 U of PRBC. There was no improvement in any ROTEM parameter during ongoing bleeding. Average values remained below the diagnostic threshold for coagulopathy (CA5 e 35 mm) throughout the bleeding episode (Fig. 2AYC) with only partial correction evident at Day 1. During trauma hemorrhage, there was profound variability in individual response to HR (Fig. 2D). Deterioration in MCF was also observed during ongoing bleeding (Fig. 3AYC). In the highest transfusion group, MCF deteriorated 15% during HR and only improved on Day 1 (Q12 U of PRBC: Time 0 MCF, 46 mm vs. Day 1 MCF, 59 mm; p G 0.05) (Fig. 3C). Again, there was significant variability to HR (Fig. 3D).
TABLE 1. Clinical Characteristics of Patients Patients
n = 106
Age, y Male Injury Injury Severity Score (ISS) ISS 9 15 Blunt injury Admission parameters Systolic blood pressure G 90 mm Hg Base deficit 9 6 mEq/L Lactate 9 5 mEq/L PT, s PTr 9 1.2 CA5 e 35, mm Damage-control procedure Operating room Interventional radiology Combined operating room/interventional radiology Fluid requirements (first 12 h) Prehospital crystalloid, mL Crystalloid, mL Colloid, mL PRBC units PRBC Q 10 U FFP units Platelet units Cryoprecipitate units
44 (30Y60) 75 (76) 34 (25Y41) 99 (94) 92 (87) 41 (46) 68 (67) 44 (42) 12.9 (11.7Y14.6) 40 (40) 41 (43) 61 (70) 6 (7) 4 (5) 500 (100Y3,000) 2,500 (1,300Y4,000) 875 (0Y1,688) 8 (5Y13) 40 (38) 6 (4Y8) 1 (1Y2) 2 (0Y2)
Values are given as n (%) or median (IQR).
Prolonged CT (994 seconds) was observed in 33 patients (31%); despite HR, this parameter also did not correct during ongoing bleeding (Fig. 4AYD), with correction only observed at Day 1. A total of 65 patients (61%) had an abnormally flattened > angle of less than 64 degrees (normal range, 63Y81 degrees) and similar to the clot strength parameters deteriorated throughout the bleeding episode (Time 0, > = 59 degrees; 4Y12 U of PRBC, > = 54Y55 degrees; Day 1, > = 70 degrees (p G 0.05).
DISCUSSION In this multicenter, prospective study of severely injured, bleeding patients, we have shown that HR does not correct hypoperfusion or coagulopathy during the acute phase of trauma hemorrhage. Transfusion of FFP/PRBC ratios greater than 1:2 throughout the bleeding episode failed to normalize any ROTEM parameters and lactic acidemia was not cleared until Day 1, after hemorrhage control had been achieved. Serial serum lactate measurements are a prognostic marker of tissue hypoperfusion and severity of hemorrhagic shock in trauma,16,19 and clearance of lactate is correlated with survival and organ failure.20 Correction of lactic acidemia remains an important therapeutic goal of HR and damage-control surgery.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
563
Khan et al.
J Trauma Acute Care Surg Volume 76, Number 3
Figure 1. Lactate clearance during hemorrhage. AYC box and whisker plots, median, IQR and adjusted range. *Versus Time 0. A, Patients receiving 4 U to 7 U of PRBC. B, Patients receiving 8 U to 11 U of PRBC. C, Patients receiving 12 U or more PRBCs.
Figure 2. Change in CA5 (mm) during bleeding episode in coagulopathic patients (CA5 e 35 mm) stratified by transfusion requirements. Dotted line is diagnostic threshold for ATC (CA5 e 35 mm). A, Patients receiving 4 U to 7 U of PRBC. Time 0: CA5, 35 mm versus Day 1: CA5, 39 mm (p G 0.05). B, Patients receiving 8 U to 11 U of PRBC. Time 0: CA5, 33 mm versus Day 1: CA5, 32 mm (p = 0.78). C, Patients receiving 12 U or more PRBCs. Time 0: CA5, 27 mm versus Day 1: CA5, 39 mm (p G 0.05). D, Individual response in all patients receiving four or more PRBC units. 564
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 76, Number 3
Khan et al.
Figure 3. Change in MCF (mm) during bleeding episode in coagulopathic patients (MCF e 54 mm) stratified by transfusion requirements. Dotted line is the diagnostic threshold for ATC (MCF e 54 mm). A, Patients receiving 4 U to 7 U of PRBC. Time 0: MCF, 55 mm versus Day 1: MCF, 58 mm (nonsignificant). B, Patients receiving 8 U to 11 U of PRBC. Time 0: MCF, 51 mm versus Day 1: MCF, 53 mm (nonsignificant). C, Patients receiving 12 U or more PRBCs. Time 0: MCF, 46 mm versus Day 1: MCF, 59 mm (p G 0.05). D, Individual response in all patients receiving four or more PRBC units.
Figure 4. Change in CT (seconds) during bleeding episode in coagulopathic patients (CT 9 94 seconds) stratified by transfusion requirements. Dotted line is diagnostic threshold for ATC (CT 9 94 seconds). A, Patients receiving 4 U to 7 U of PRBC. Time 0: CT, 108 seconds versus Day 1: CT, 69 seconds (p G 0.05). B, Patients receiving 8 U to 11 U of PRBC. Time 0: CT, 107 seconds versus Day 1: CT, 69 seconds (p = 0.07). C, Patients receiving 12 U or more PRBCs. Time 0: CT, 126 seconds versus Day 1: CT, 63 seconds (p G 0.05). D, Individual response in all patients receiving four or more PRBC units. * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
565
J Trauma Acute Care Surg Volume 76, Number 3
Khan et al.
Combination of these strategies in combat casualties has been shown to correct metabolic derangements in a before and after study of hemorrhage control.10 Our study is consistent with this, but we have shown that restoration of normal tissue perfusion and improvement in TIC does not occur until hemorrhage control is achieved. Clot strength and CTs deteriorated during ongoing bleeding and only corrected when no further PRBC units were required. Our findings reinforce those from a comparative study of laboratory-guided MHP versus empiric 1:1:1 (PRBC/FFP/ platelets) protocol.13 Fibrinogen, platelets, and prothrombin time trends in the acute bleeding episode were identical in both transfusion strategies, despite a fourfold increase in patients receiving high FFP/PRBC ratio (91:2). HR may offer several advantages over historical transfusion strategies, for example, rapid access to empiric component therapy, but the findings from this study show it has limited efficacy in correcting TIC. The benefits of HR have been postulated to a reduction in CSL use. Studies have shown that increased CSL use is associated with dilutional coagulopathy and worse outcomes.9 Permissive hypotension was practiced at all study sites with limited volumes of CSL infused before baseline blood samples. More severe changes in coagulopathy may have been seen had HR protocols not been followed by centers in this study. HR may therefore protect somewhat from deterioration of these parameters during bleeding but in itself is not able to correct TIC. There are several limitations to this study. We were only able to sample up to the 12th PRBC unit transfused and therefore could not follow coagulation profile trends up to the point of hemorrhage control in some patients. Precise timing of hemorrhage control is difficult to ascertain, and therefore, we used completion of PRBC transfusion as a surrogate measure for control of bleeding. Average FFP/PRBC ratios were greater than 1:2 throughout, and we did not look specifically at the effect of differential ratios on coagulation or hypoperfusion. We also did not look at other component therapies (platelets and cryoprecipitate), although clearly, they will influence correction of TIC. In particular, fibrinogen is fundamental to hemostasis and falls to critical levels soon after the onset of major trauma hemorrhage.11 Early aggressive fibrinogen replacement may therefore prove efficacious in the correction of TIC and merits further investigation. We did not examine the fibrinolytic component of TIC in this study, although from 2010 and the publication of the CRASH-2 trial, MHPs at all sites specified early administration of tranexamic acid. Undertreated or persistent hyperfibrinolysis may contribute to poor clot function and is therefore part of continued investigation within INTRN. This investigation questions the therapeutic mechanism and efficacy of HR in the acute phase of trauma hemorrhage. Aggressive high-volume replacement of plasma has little effect on deranged ROTEM coagulation parameters and hypoperfusion. Arrest of hemorrhage and cessation of transfusion seem necessary before TIC can be corrected and tissue perfusion restored. Following hemorrhage control and completion of transfusion, ROTEM parameters do not immediately return to values seen in normal healthy controls. The question for future design of goal-directed therapy, using devices such as ROTEM remains the identification of target values for coagulation 566
parameters in trauma hemorrhage. Significant opportunities exist to tailor management and improve outcomes for bleeding trauma patients. A balanced and potentially novel strategy is required to achieve correction of TIC while maintaining organ perfusion. HR as practiced by the institutions in this study may be protective but is neither hemostatic nor resuscitative. AUTHORSHIP S.K., K.B., and R.D. conceived the study. R.D., S.K., I.R., and M.C. collected and analyzed the data. R.D., C.G., S.S., and K.B. wrote and edited the manuscript.
DISCLOSURE TEM Innovations (ROTEM) provided unrestricted support in the form of equipment and reagent support for ACIT study. S.K. and R.D. have received honorarium as invited speakers. This study was funded in part by the National Institute for Health Research (UK) Program Grant for Applied Research (RP-PG-0407-10036).
REFERENCES 1. Holcomb JB, Del Junco DJ, Fox EE, Wade CE, Cohen MJ, Schreiber MA, Alarcon LH, Bai Y, Brasel KJ, Bulger EM, et al. The Prospective, Observational, Multicenter, Major Trauma Transfusion (PROMMTT) study: comparative effectiveness of a time-varying treatment with competing risks. JAMA Surg. 2013;148(2):127Y136. 2. Mitra B, Cameron PA, Mori A, Fitzgerald M. Acute coagulopathy and early deaths post major trauma. Injury. 2012;43(1):22Y25. 3. Brohi K, Cohen MJ, Davenport RA. Acute coagulopathy of trauma: mechanism, identification and effect. Curr Opin Crit Care. 2007;13(6):680Y685. 4. Gonzalez EA, Moore FA, Holcomb JB, Miller CC, Kozar RA, Todd SR, Cocanour CS, Balldin BC, McKinley BA. Fresh frozen plasma should be given earlier to patients requiring massive transfusion. J Trauma. 2007; 62(1):112Y119. 5. Murad MH, Stubbs JR, Gandhi MJ, Wang AT, Paul A, Erwin PJ, Montori VM, Roback JD. The effect of plasma transfusion on morbidity and mortality: a systematic review and meta-analysis. Transfusion. 2010; 50(6):1370Y1383. 6. Ho AM, Karmakar MK, Dion PW. Are we giving enough coagulation factors during major trauma resuscitation? Am J Surg. 2005;190(3):479Y484. 7. Holcomb JB, Jenkins D, Rhee P, Johannigman J, Mahoney P, Mehta S, Cox ED, Gehrke MJ, Beilman GJ, Schreiber M, et al. Damage control resuscitation: directly addressing the early coagulopathy of trauma. J Trauma. 2007;62(2):307Y310. 8. Brohi K, Singh J, Heron M, Coats T. Acute traumatic coagulopathy. J Trauma. 2003;54(6):1127Y1130. 9. Maegele M, Lefering R, Yucel N, Tjardes T, Rixen D, Paffrath T, et al. Early coagulopathy in multiple injury: an analysis from the German Trauma Registry on 8724 patients. Injury. 2007;38(3):298Y304. 10. Morrison JJ, Ross JD, Poon H, Midwinter MJ, Jansen JO. Intra-operative correction of acidosis, coagulopathy and hypothermia in combat casualties with severe haemorrhagic shock. Anaesthesia. 2013;68(8):846Y850. 11. Rourke C, Curry N, Khan S, Taylor R, Raza I, Davenport R, Stanworth S, Brohi K. Fibrinogen levels during trauma hemorrhage, response to replacement therapy, and association with patient outcomes. J Thromb Haemost. 2012;10(7):1342Y1351. 12. Davenport R, Curry N, Manson J, De’Ath H, Coates A, Rourke C, Pearse R, Stanworth S, Brohi K. Hemostatic effects of fresh frozen plasma may be maximal at red cell ratios of 1:2. J Trauma. 2011;70(1):90Y95; discussion 95Y96. 13. Chambers LA, Chow SJ, Shaffer LE. Frequency and characteristics of coagulopathy in trauma patients treated with a low- or high-plasma-content massive transfusion protocol. Am J Clin Pathol. 2011;136(3):364Y370.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 76, Number 3
Khan et al.
14. Ganter MT, Hofer CK. Coagulation monitoring: current techniques and clinical use of viscoelastic point-of-care coagulation devices. Anesth Analg. 2008;106(5):1366Y1375. 15. Davenport R, Manson J, De’Ath H, Platton S, Coates A, Allard S, Hart D, Pearse R, Pasi KJ, MacCallum P, et al. Functional definition and characterization of acute traumatic coagulopathy. Crit Care Med. 2011;39(12): 2652Y2658. 16. Re´gnier MA, Raux M, Le Manach Y, Asencio Y, Gaillard J, Devilliers C, Langeron O, Riou B. Prognostic significance of blood lactate and lactate clearance in trauma patients. Anesthesiology. 2012;117(6): 1276Y1288. 17. Husain FA, Martin MJ, Mullenix PS, Steele SR, Elliott DC. Serum lactate and base deficit as predictors of mortality and morbidity. Am J Surg. 2003;185(5):485Y491. 18. Abramson D, Scalea TM, Hitchcock R, Trooskin SZ, Henry SM, Greenspan J. Lactate clearance and survival following injury. J Trauma. 1993;35(4):584Y588; discussion 588Y589. 19. Odom SR, Howell MD, Silva GS, Nielsen VM, Gupta A, Shapiro NI, Talmor D. Lactate clearance as a predictor of mortality in trauma patients. J Trauma Acute Care Surg. 2013;74(4):999Y1004. 20. Manikis P, Jankowski S, Zhang H, Kahn RJ, Vincent JL. Correlation of serial blood lactate levels to organ failure and mortality after trauma. Am J Emerg Med. 1995;13(6):619Y622.
DISCUSSION Dr. Yashuhiro Otomo (Tokyo, Japan): The combination of fibrinolysis dilution and inadequate function produces a complex coagulopathy in many trauma patients who receive massive transfusions. An aggressive transfusion therapy with plasma and platelet could prevent or correct deficiencies and, thereby, might improve survival. Indeed, multiple trauma programs have achieved dramatically increased survival rate. However, most of these studies are retrospective, case controlled review of notoriously complex patient situations with many injuries and treatment features that affect outcome. Dr. Davenport and coauthors have conducted this prospective cohort study of bleeding trauma patients presenting directly to three major international trauma centers. This study was well designed and organized and successfully collected a large amount data at each time point. The authors may predict a positive effect of aggressive transfusion of plasma on the outcome and also aimed to describe the effect of varying FFP:RBC ratios on correction of coagulopathy. However, surprisingly, in this multi-center prospective study hemostatic resuscitation was not effective to correct hypoperfusion or coagulopathy during the acute phase of trauma hemorrhage. From these results the authors made this provocative title: Hemostatic Resuscitation is Neither Hemostatic Nor Resuscitative in Trauma Hemorrhage. This is the first prospective study to show an evidence that aggressive transfusion of plasma have very few effects in reversing coagulopathy caused by severe trauma bleeding. I want to commend the excellent study group for this great achievement. I have several comments and questions for the authors. First, the definition of hemostatic resuscitation is not clear. My understanding of hemostatic resuscitation is the combination of avoiding dilution and hypothermia, enough amount of administration of FFP, administration of fibrinogen, PCC or cryoprecipitate and administration of anti-hypofibrinolytic agent.
In your study you should clearly describe the definition of hemostatic resuscitation. Second, among the total 106 patients that you included into your analysis you mentioned overall 69that is, 55% of patientswere coagulopathic by reduced clot strength, CA5 less than 36 millimeters. This means 35% of patients included in your analysis are not coagulopathic. I believe that this study should include only coagulopathic patients or you should show the data analyzed only on the 35%non-coagulopathic patients. Third, in your study, the timing that hemorrhage control was achieved is not clearly described. From your presentation that timing appears to be the Day 1; however, I assume that that timing would be different from case to case, such as within two hours of the injury or 24 hours after injury. Please make it clear. Fourth, you didn’t mention about the survival bias. As you know, the clinical researches on transfusion ratios, one of the most important issue that should be addressed has been this survival bias. In order to evaluate the efficacy of aggressive FFP transfusion you should touch upon it. Finally, a massive transfusion of FFP provides a massive amount of protein C to the patient. According to your theory this would cause a large amount of production of activated protein C that will result in a progression of trauma-induced coagulopathy. Why did you predict that high ratio of FFP transfusion might be effective for trauma massive hemorrhage? That is competing against your theory of trauma-induced coagulopathy. The authors said a balanced and potentially novel strategy is required. I completely agree with that. I would like to thank the association for the privilege of the floor. Dr. John B. Holcomb (Houston, Texas): Dr. Davenport, I congratulate you on a very clever title. Dr. Brohi’s favorite term, I think. The problem with this paper is it is observational. And so deriving over reaching conclusions with no control group greatly overstates your data. Importantly, you nicely presented a lot of laboratory data but you didn’t tell us how the patients did. Probably the most important missing element here is time. All of this interventions are extraordinarily timedependent. When did the patients die? What did they die from. How did they do? I congratulate you for partially adopting the hemostatic or damage control methods by limiting crystalloid and then trying to achieve an early ratio. But in fact you didn’t. You didn’t start plasma early. You caught up late. The PROMMTT trial of 1200 patients recently published several papers in the Journal of Trauma. One of them directly addressed this issue: lack of plasma early, catching up later, resulted in poor outcomes. Please comment. Dr. Richard J. Mullins (Portland, Oregon): This study raises the question: is there surgeon variation that accounts for your outcome? Your study shows that massive transfusion has associated with a variation in death but how have you accounted for the fact that the patients who had four units of blood were operated on by a master surgeon like Rao Ivatury and the patients who got 12 units of blood were operated on by somebody who occasionally takes trauma call?
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
567
J Trauma Acute Care Surg Volume 76, Number 3
Khan et al.
The general principle is when you are going to talk about bleeding to death I don’t know how you are going to account for the fact that there is in fact variation in the ability of surgeons to stop bleeding that has a major bias in your data. Dr. Eileen M. Bulger (Seattle, Washington): I think the fundamental question you are asking is unfair. I mean I think there is no question that we have to control bleeding to correct coagulopathy and if you can’t control bleeding your patient is going to die. So as much resuscitation as we do, we’re trying to keep up and not to completely correct coagulopathy during the time of bleeding. I think that’s an unfair goal. I also want to echo Dr. Holcomb’s concerns about timing and survival bias because absolutely dividing patients by the amount of blood they got, if they die early and they don’t get a lot of blood, it’s a very mixed group in that low blood use group so you have to account for that in the analysis. And I really want to know how you accounted for early deaths in this study and how you accounted for these timing issues of people who achieve hemostasis at different times and people who get ratios at different times. Dr. Mitchell Jay Cohen (San Francisco, California): I echo the comments that have been previously stated. I think this paper begins to address one of the central issues in hemostatic resuscitation which is whether these patients do better from hemostatic resuscitation because they stop bleeding and don’t bleed to death or do they do better because there is something else in the ‘‘special sauce’’ that is keeping the patients and, their endothelium happy and keeping them alive? So my question to you is what else is in your data? First of all, did these patients actually have better outcomes? What does the outcome data look like when they got hemostatic resuscitation? And if it’s not bleeding and it’s not lactate clearance, what else is going on, in your opinion? Dr. Zsolt Balogh (Newcastle, Australia): Congratulations, especially supporting my bias on eliminating crystalloids is probably the cause for better outcomes. I must highlight that damage control resuscitation was introduced based on much lower quality studies than this one, so I don’t think it is too bad. My question is how long was the time frame of hypotension of these patients and what time you managed to catch up with the desired goal of blood and FFP during the resuscitation? Dr. Louis Alarcon (Pittsburgh, Pennsylvania): Two questions, please. First, did you include tranexamic acid in your massive transfusion protocol? And the second question relates to the PROMMTT study, wherein we learned that the ratios of blood products varied over time in any particular patient. So my questions are: at what time point did you calculate the plasma and platelet to RBC ratios? And how did you account for the fact that these ratios, in any particular patient, vary considerably over time? Thank you. Dr. Ross Davenport (London, United Kingdom): Thank you all for those insightful comments. Clearly the title was
568
provocative and is designed to generate discussion and to consider alternative hypotheses. Firstly, to answer some of Professor Otomo’s comments. All the centers adhered to the principles of hemostatic resuscitation. Fibrinogen supplementation was administered in the form of cryoprecipitate in addition to platelets. All centers now use tranexamic acid. Some of the patients were involved in this study prior to the publication of the CRASH-2 trial, therefore, not all of the patients received TXA although the majority did. We only included coagulopathic patients because we felt that there was limited benefit in examining those who are not coagulopathic as FFP would not be expected to alter their coagulation profiles. The timing of hemorrhage control is an important and vexing question and the comments made by Dr. Holcomb are very important. Clearly, the more shocked a patient is or the longer they remain shocked the worse the coagulopathy is likely to be and further drive the bleeding, shock and coagulopathic process. All patients in this study had damage control resuscitation by attendings who only take trauma calls. It is difficult to ascertain the precise time to hemorrhage control. However all these patients had a massive hemorrhage protocol would have had control of hemorrhage within the first one-to-two hours of arrival in the emergency department. We didn’t look specifically at the effects of different ratios in this study. We have previously shown that differential ratios of FFP to packed red cells have minimum effect on the coagulation profile. The maximum benefit is seen at a ratio of one to two. Clearly, this is something that does require further examination. And I think we have to think of newer ways to actually work out what rapid bleeding means and how to precisely evaluate the effects of other component therapies, clearly, platelets and fibrinogen supplementation. You asked for my comments on tranexamic acid. It’s always a topic of controversy when we come to these sorts of meetings. As I’ve said, the patients included in this study were both before and after 2010 (CRASH-2). There are clearly many questions that still need to be answered regarding tranexamic acid, certainly how it actually works, the dosing, optimal timing, and its precise role in a mature trauma system. Lastly, the comments about activated protein C. We have shown that this is a potential mediator of this coagulopathy. Those patients that received high doses of plasma could theoretically, further drive generation of activated protein C and coagulopathy. Regarding limitation of crystalloids and actually what FFP does, I think this is an important question. There are numerous factors in FFP which we don’t know, what are their effects on coagulation or inflammation and whether or not it acts early or late, for example in stabilization of the endothelial glycocalyx. Thank you.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
