| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
*Institute for Anesthesiology and
Institute for Clinical Chemistry and Hematology, Kantonsspital, St. Gallen, Switzerland; and
Department of Clinical Research, University of Bern, Bern, Switzerland
Address correspondence and reprint requests to Wolfgang Korte, MD, Institute for Clinical Chemistry and Hematology, Kantonsspital, CH 9007 St. Gallen, Switzerland. Address e-mail to wolfgang.korte{at}ikch.ch
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
0.014). However, F. XIII availability per unit thrombin generated was significantly decreased in bleeders before, during, and after surgery (P 0.051). Computerized thrombelastography showed a parallel, significant reduction in clot firmness. We suggest that mild preexisting coagulopathy is not rare in surgical patients and probably can result in clinically relevant intraoperative bleeding. This hemostatic disorder shows impaired clot firmness, probably secondary to decreased cross-linking (due to a loss of F. XIII, both in absolute measures and per unit thrombin generated). We suggest that the application of F. XIII might be worthwhile to test in a prospective clinical trial to increase clot firmness in patients at risk for this intraoperative coagulopathy. IMPLICATIONS: From a prospective study of 226 patients undergoing elective surgery, we suggest that unexplained intraoperative bleeding might occur secondary to a preexisting, nonovert coagulopathy with decreased Factor XIII availability and a consecutive reduction of clot firmness during surgical stress.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Besides the patient’s history and clinical assessment (4,5), alterations of various laboratory variables, including prothrombin time (PT), activated partial thromboplastin time (aPTT), and platelet count, may be associated with unexpected bleeding (6). However, their use as routine screening assays is controversial (5,7,8) because blood loss can be unexpectedly large despite normal preoperative values (7,9). Thus, either these assays are not sensitive enough to detect a coagulopathy that leads to intraoperative, unexplained bleeding, or the reason for the coagulopathy develops only during surgery and can therefore not be detected beforehand. We have previously reported (10) that patients with unexplained intraoperative bleeding show increased preoperative coagulation activation: we found these patients to have significantly higher preoperative thrombin generation (prothrombin fragments F1 + 2) and fibrin degradation (D-dimer). Thus, given that these patients showed unexplained intraoperative bleeding despite increased coagulation activation, we hypothesized that bleeding might result from problems of fibrin cross-linking. We therefore attempted to further elucidate this phenomenon through the evaluation of elements of the last steps of the coagulation cascade leading to fibrin formation. We evaluated markers of coagulation activation, fibrin generation, and fibrin cross-linking before, during, and after surgery in patients undergoing elective surgical procedures. The goal of this study was to identify elements in the final pathway of the coagulation cascade that might be responsible for diffuse and unexplained intraoperative bleeding.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Inclusion and exclusion criteria were as follows: (a) surgery had to start and terminate during the normal day shift of the clinical laboratory; (b) an arterial catheter had to be inserted before surgery to guarantee adequate blood sampling without artifacts (the decision for its insertion was based on clinical criteria only, which included, but were not restricted to, significantly increased circulatory and/or cardiac risk, anticipated risk for extensive bleeding, or planned combined general and [thoracic] epidural anesthesia); (c) no patients on extracorporal circulation were to be included; and (d) no patients undergoing intentional hypothermia were to be included. No other exclusion criteria were set; however, other variables such as abnormal preoperative hemoglobin concentration, platelet count, PT ratio, or aPTT; the use of medications such as anticoagulants or platelet activation inhibitors; and a history of easy bruising or coagulopathy were recorded and evaluated.
Biometric data, body mass index (BMI), and ASA physical status, as well as data concerning the course of the surgical procedure (duration, blood loss, infusion volumes, quantity and type of administered blood products, and so on), were prospectively collected with a predefined datasheet. The study was planned to include markers of thrombin and fibrin generation; the decision to evaluate Factor XIII (F. XIII) activity and clot firmness was made post hoc.
For infusion (crystalloids and plasma expanders) and transfusion (packed red blood cells, fresh frozen plasma (FFP), and platelet concentrates) therapy, specific guidelines for the calculation of critical threshold values were to be followed; treatment goals were defined according to the patient’s age and cardiovascular risk factors. After infusion of at least 1000 mL of crystalloids, plasma expanders were used to substitute for blood loss only when transfusion triggers were reached (see below). Gelatin solution (GS) (Physiogel®) and hydroxyethyl starch (HS) (HAES® 6%) were allowed; however, the administration of HS was strictly limited to a maximum of 30 mL/kg. Clinical threshold values for the administration of packed red cells were hemoglobin levels <70 or 100 g/L (depending on the situation and the cardiovascular morbidity of the patient); FFP was used for bleeding with prothrombin ratio <0.5; and platelet transfusions were given for bleeding with platelet counts <50 or 100 g/L (depending on the clinical situation and the type of surgery performed). Decisions to use vasopressors were left to the anesthesiologist in charge of the patient.
The defining properties of nonmechanically induced (i.e., unexplained) intraoperative bleeding were fixed preemptively and before the study was started. The occurrence of unexplained intraoperative bleeding according to these criteria was evaluated prospectively. Because there is no satisfactory approach to quantify the unexplained bleeding fraction during a surgical procedure, we chose a purely clinically based definition, in accordance with other investigators (11,12). The presence of unexplained bleeding was assessed by the anesthesiologist in charge of the patient and was defined by (a) normal, clinically undisturbed, and adequate local and systemic hemostasis followed by (b) de novo occurrence of unexplained bleeding from wound margins or from puncture sites; (c) a diffuse, nonlocalized bleeding type; and (d) no vessel stumps identified as potential sources of bleeding. For quantification of the bleeding, blood loss was calculated from the measured blood volume drained from the surgical field and the estimated volume in surgical sponges. Patients with unexplained intraoperative bleeding are referred to as "bleeders" and those without as "nonbleeders."
All blood samples were taken from the arterial catheter. The first 3 mL was used for nonstudy purposes; thereafter, 3.6 mL was drawn into 0.125 mol/L sodium citrate (0.4 mL) by using a vacuum system (Vacutainer; Becton Dickinson). The arterial catheter was flushed with 0.9% sodium chloride solution only: no anticoagulant was used at any time. Samples were taken at five different time points:
Blood samples were directly sent to the laboratory via pneumatic dispatch system. Platelet-poor plasma was obtained by centrifugation (10 min, 1600g, 22°C), aliquoted into plastic tubes, and frozen and stored at –80°C until used. Study data were not accessible to anyone in the operating rooms or intensive care units.
Assays were performed according to the recommendations of the manufacturers. They included prothrombin fragment F1 and 2 (reference range, 0.4–1.1 nmol/L; Enzygnost F1 + 2 micro; Dade Behring, Marburg, Germany), fibrinogen (reference range, 1.5–3.5 g/L; Clauss method, run on a BCS analyzer; Dade Behring), fibrin monomer (reference range, <9 µg/L; Enzymun FM, run on an ES300; Roche Diagnostics, Mannheim, Germany), and F. XIII activity (reference range, 70%–140%; Berichrome [Dade Behring] on a BCS analyzer). Computerized plasma thrombelastography (ROTEM® coagulation analyzer; Pentapharm Ltd., Basel, Switzerland) was used (13,14) to quantify maximum clot firmness (MCF) in order to evaluate changes in clot quality over time in the group of bleeders (n = 20) and a matched sample of nonbleeder patients (matched for type and duration of surgery, sex, age, BMI, and ASA physical status).
Data were compared between bleeders and nonbleeders by using the Mann-Whitney ranked sum test (biometric data, infusion volumes, blood loss, and blood product support; all coagulation variables and thrombelastographic data shown). Chi-square testing was used to compare frequencies between bleeder and nonbleeder groups. Statistical data were analyzed with SigmaStat Version 3.0. Data were logarithmized to normalize data and allow plotting of mean and SEM error bars (SigmaPlot Version 8.02; all SPSS, Zürich, Switzerland).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Predefined clinical criteria (see Methods) were used to assign patients to the bleeder (n = 20) or nonbleeder (n = 206) groups. A comparison of biometric data between groups is shown in Table 1. A history of spontaneous hematoma and/or easy bruising was not more frequent in the bleeder than the nonbleeder group (0% versus 0.5%; 
2 = 1.77; not significant). The same was true for preoperative ingestion of drugs inhibiting platelet activation (acetyl salicylic acid, nonsteroidal antiinflammatory drugs, and similar; 5% versus 5.5%; 2 = 0.171; not significant) and preoperative heparin administration (45% versus 56%; 2 = 0.48; not significant). Preoperative thrombocytopenia was usually mild and similarly frequent in the bleeder and the nonbleeder group (16% versus 6%; 2 = 1.23; not significant); the same was true for an abnormal PT ratio (5% versus 2.5%; 2 = 0.002; not significant) and an abnormal aPTT (5% versus 2%; 2 = 0.008; not significant).
|
2 < 0.001; not significant). There were also no differences in HS application between the bleeder and the nonbleeder groups (median [range], 0 mL [0–1500 mL] versus 0 mL [0–2000 mL], respectively; not significant). However, significantly more GS was given to bleeders (1000 mL [0–4500 mL]) than nonbleeders (0 mL [0–3000 mL]; P < 0.001). In addition, the lowest body temperature measured was lower in bleeders than nonbleeders (33.9°C [32.1°C–35.5°C] versus 34.7°C [32.8°C–38.3°C]; P = 0.003). Differences in median intraoperative blood loss between bleeders (1350 mL) and nonbleeders (400 mL; P < 0.001) became apparent at T3, whereas no significant difference was seen at T2 (Table 1). The need for replacement with blood products was significantly increased in the bleeder group (P < 0.001 for packed red cells at T3–4 and for FFP at T4; Table 1).
Fibrin generation increased during surgery, and patients who developed unexplained intraoperative bleeding had higher fibrin generation before, during, and after surgery compared with nonbleeders (Table 2). Fibrinogen decreased significantly in bleeders during surgery by T3 (Table 2), as did F. XIII activity (Fig. 1). The amount of F. XIII (activity) available per unit thrombin was decreased at any time (Fig. 2), albeit with borderline statistical significance at T2 (Table 2). In addition, mean clot firmness as assessed by computerized thromboelastography was significantly less in bleeders than nonbleeders when compared with baseline values (Table 2). This difference could already be detected at T2, whereas bleeding volume and blood product support were not yet different between groups (Fig. 3).
|
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Low F. XIII levels have repeatedly been implicated in intraoperative bleeding complications and/or increased requirement for blood products (17–21), and major surgery is one of several causes of acquired F. XIII deficiency (22). According to the reduced availability of F. XIII per unit thrombin generated and the significant decrease of F. XIII activity in bleeders, it seems reasonable to postulate that early substitution of F. XIII might reduce the loss of clot firmness and thus finally reduce bleeding (23–25) in patients at risk. Bleeders received significantly more GS than nonbleeders, had significantly lower body temperatures, and were operated on for a longer time. All these properties could theoretically contribute to a bleeding tendency; however, clot firmness was already significantly reduced in the bleeder group at T2 (a fixed time point 30 minutes into surgery) even before clinical bleeding became apparent in this group. In addition, the infusion triggers used prevented the use of plasma expanders in the absence of significant bleeding, thus suggesting that the use of plasma expanders was secondary to bleeding problems rather being the reason for the increased bleeding. Because we did not measure body temperature at the very moment when the unexplained bleeding tendency occurred and used the lowest reading recorded for evaluation, we cannot decide whether the lower body temperature was the consequence of or the reason for the bleeding in the bleeder group, but it seems reasonable to believe that the lower temperature was a consequence of, rather than the reason for, the bleeding, given the longer surgery time in these patients.
Our study has some limitations; notably, patients with different diseases and types of surgery were included. However, in a subgroup analysis (evaluating patients with cancer only), matched comparison of bleeders and nonbleeders revealed similar results for hemostatic variables compared with the results of the entire group (data not shown). Another point is the open study design with regard to the exclusion criteria, such as preoperative coagulation status and medications. However, again, no significant differences were observed between groups in the frequencies of these factors.
In conclusion, our results suggest that unexplained intraoperative bleeding (as defined and described above) may develop from of a preexisting, but preoperatively clinically silent, activation of the coagulation system. Upon surgical stress, features of consumptive coagulopathy occur that also seem to influence cross-linking of fibrin. A significant loss of clot firmness develops that can be detected early during surgery, even before bleeding becomes clinically apparent. Initial experience with early F. XIII substitution suggests that this approach merits further evaluation in a prospective clinical trial.
|
Acknowledgments |
|---|
We would like to thank all personnel involved in this study (especially laboratory technicians and anesthesiology nursing staff), as well as our colleagues in the departments of surgery, orthopedics, neurosurgery, and urology.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. A. Tanaka, N. S. Key, and J. H. Levy Blood Coagulation: Hemostasis and Thrombin Regulation Anesth. Analg., May 1, 2009; 108(5): 1433 - 1446. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Haas, D. Fries, C. Velik-Salchner, C. Reif, A. Klingler, and P. Innerhofer The In Vitro Effects of Fibrinogen Concentrate, Factor XIII and Fresh Frozen Plasma on Impaired Clot Formation After 60% Dilution Anesth. Analg., May 1, 2008; 106(5): 1360 - 1365. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. T. Ganter and C. K. Hofer Coagulation Monitoring: Current Techniques and Clinical Use of Viscoelastic Point-of-Care Coagulation Devices Anesth. Analg., May 1, 2008; 106(5): 1366 - 1375. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Mittermayr, W. Streif, T. Haas, D. Fries, C. Velik-Salchner, A. Klingler, E. Oswald, C. Bach, M. Schnapka-Koepf, and P. Innerhofer Hemostatic Changes After Crystalloid or Colloid Fluid Administration During Major Orthopedic Surgery: The Role of Fibrinogen Administration Anesth. Analg., October 1, 2007; 105(4): 905 - 917. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
