| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
*Department of Anesthesiology, Duke University Medical Center, Durham, North Carolina; Department of Anaesthesia, University of Cape Town, Cape Town, South Africa; and Centre for Anaesthesia, University College London, Middlesex Hospital, London, United Kingdom
Address correspondence to Anthony M. Roche, Department of Anesthesiology, Duke University Medical Center, DUMC 3094, Durham, NC 27710. Address e-mail to tony.roche{at}duke.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Some hydroxyethyl starch (HES) solutions have been associated with clinical coagulopathies and bleeding when administered IV in large volumes (4). Highly substituted, large molecular weight (MW) starches (5,6), with high C2:C6 ratios (7) are more likely to induce coagulation derangements than smaller MW, lesser substituted starches with lower C2:C6 ratios. Furthermore, it has been suggested that colloids suspended in balanced salt (BAL) solutions are associated with better coagulation profiles than those based in saline (8,9). Previous work by our group has demonstrated differences in coagulation (10), bleeding (8), acid-base profile, and gastrointestinal tract function (11) between starches in different carrier solutions.
The thromboelastograph® (TEG®) is a point-of-care device that is being used in many clinical environments. It analyzes the real-time ability and rate of the blood coagulation. The trace obtained is due to the oscillating action of the piston in the cup, transmitting a signal to the output. This trace is then analyzed by various times, measurements and angles of the trace. The r time, K time, 
angle, and maximum amplitude (MA) are standard measures of the device. The r time reflects the time to onset of clot formation, the K time and angle reflect the rate of clot formation, and the MA reflects the total strength of developing clots. Traces with prolonged r and K times and decreased angles and MA are associated with a hypocoagulability; decreased r and K times and increased angles and MA characterize a hypercoagulable state (12).
No published studies have examined coagulation function between two BAL salt HES compounds and similar saline-based products at a wide-range of dilutional levels. We therefore conducted an in vitro study investigating the coagulation effects of a number of different crystalloid and colloid IV fluid preparations, as judged by the TEG® (Haemoscope Corp, Chicago, IL). Our hypothesis was that similar starch compounds suspended in BAL salt carriers maintain coagulation variables better than those in saline-based carrier solutions.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Fresh blood was diluted by gentle inversion eight times each in polypropylene containers with the following fluids in random order to dilutions of 20%, 40%, and 60%:
Diluted samples, as well as fresh undiluted blood controls, were placed in calibrated TEG® analyzers and analysis was begun 6 min from venipuncture. The TEG® channels used were also changed to avoid systemic errors caused by small differences in the individual channels. Assays were performed according to the manufacturer's guidelines (Haemoscope Corp.). All samples not producing r-times by 60 min were deemed to have failed to clot. Prior agreement of the investigators was that unclotted samples were assigned r- and K-times of 60 min each, with alpha angles and MA assigned values of 0. HAS 4.5% produced flat line TEG® traces in all samples at 60% dilution (i.e., no evidence of clot formation), therefore no further dilutions of 60% HAS 4.5% were performed after the first five volunteers.
Ionized calcium levels were analyzed in all samples using a Critical Care Laboratory Synthesis 25 blood gas and electrolyte machine (Instrumentation Laboratory, Lexington, MA). This device underwent daily quality control, twice-daily two-point calibrations, and single-point calibrations after each processed sample.
All data were analyzed by analysis of variance, with post hoc least significant difference analysis performed where significant differences were detected by analysis of variance. These were performed on STATISTICA® Version 6 software (StatSoft Inc, Tulsa, OK), running on a Microsoft Windows®-based IBM®-compatible personal computer. P values of less than or equal to 0.05 were considered significant. Based on previous work from our own laboratory, as well as other published data, 10 volunteers were estimated to adequately power the study (80%) at the 5% significance level to detect 20% differences in TEG® variables from control samples.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
angles (Fig. 2), and MA (Fig. 3) have been displayed. All data are expressed as mean ± sd. Note that control values in the graphs represent undiluted fresh whole blood samples analyzed at the same time as the diluted samples.
|
|
|
At 20% dilution, no differences from control were found with the r-times of any fluids. Tetr-NaCl displayed reduced MA values compared with undiluted controls. At 40% dilution, HES-NaCl was the only fluid to display prolonged r-times compared with control. Also at 40% dilution, HES-NaCl, HES-Bal, Tetr-NaCl, and Alb all displayed reduced MA values compared with control. At 60% dilution, NaCl, HES-NaCl, HES-Bal, Pent-NaCl, Tetr-NaCl, and Alb all displayed prolonged r-times compared with undiluted controls, whereas all fluids exhibited reduced MA values compared with controls.
The crystalloids both enhanced coagulation, as measured by the angle, at 40% dilution, as did the 200/0.5 starch in a BAL salt solution (Pent-Bal). At 60% dilution, all of the colloids impaired coagulation, except for the 200/0.5 starch in a BAL solution (Pent-Bal).
The ionized calcium levels (Fig. 4) of the saline-based fluids were significantly lower than controls, especially at 40%–60% dilutions. At 20% dilutions, ionized calcium concentration was above the critical level of 0.56 mmol/L at which calcium affects coagulation (13). At 40% dilution, only HAS resulted in calcium concentrations below this critical value, but at 60% dilution, all of the saline-based fluids approached or were smaller than this concentration.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In a study of human subjects who received pentastarch during leukapheresis, Strauss et al. (5) reported that the medium MW pentastarches caused less derangement of coagulation indices than large MW hetastarches. Treib et al. (6) reported that not only the SR but also the C2:C6 ratio (7) was of importance in causing a hypocoagulable state. This is due to the highly substituted (more glucose rings undergoing hydroxyethyl substitution), high C2:C6 ratio (relatively more hydroxyethyl groups in the C2 than C6 glucose positions) starches undergoing slower metabolism (i.e., translating into a larger mean MW of the starch in vivo). It is interesting that the saline-based 450/0.6 starch (HES-NaCl) in our study produced significantly more intense impairment of coagulation than any other solution except Alb. Indeed, these Alb results may be surprising, because it has not commonly been reported to cause a hypocoagulable state. Tobias et al. (14), in an in vitro study of healthy volunteers, reported that Alb led to a significantly hypocoagulable state. On the other hand, Vogt et al. (15,16), Niemi and Kuitunen (17), and Kuitenen et al. (18) separately concluded that Alb caused minimal coagulation derangement in their in vivo clinical studies of Alb compared with HES preparations. Alb binds calcium more so than the other fluids we studied, which can be seen in the calcium results reported in Figure 4. In the clinical setting, if Alb is administered in large quantities, a hypocalcemic state may ensue. It may be prudent to monitor and supplement calcium appropriately, thereby potentially minimizing significant coagulation effects; however, this has not been rigorously studied.
The smaller MW pentastarches and tetrastarches have been commercially available in Europe for more than a decade, and these products are used more often than fluids containing large MW hetastarches. Tetr-NaCl has been approved for clinical use in Europe for approximately 5 years. In the United States (USA), only two starch solutions are currently clinically available for intravascular volume resuscitation, a 6% large MW hetastarch, suspended in saline (e.g., Hespan®) and 6% hetastarch suspended in a BAL salt solution, Hextend® (HES-Bal, used in this study). At the time of writing, a medium MW starch (Pentaspan®; B. Braun, Bethlehem, PA) is approved by the Food and Drug Administration in the USA for plasmapheresis but not for intravascular volume resuscitation or expansion. Pent-Bal is currently undergoing Phase II investigation in the USA.
Calcium concentrations at the various dilutions may account for the differences observed, especially at 60% dilution. We have shown previously that an ionized calcium concentration of more than 0.56 mmol/L is required for normal coagulation function as judged by TEG® (13). This is most likely to have been of greatest influence in the coagulation function we observed with Alb, because Alb binds calcium. Ionized calcium levels for Alb already decreased to 0.44 and 0.26 mmol/L at 40% and 60% dilutions, respectively. Most of the TEG® effects observed for Alb at these dilutions were probably due to hypocalcemia. A similar argument can be made for the saline-based fluids, especially at 60% dilutions, because most of the ionized calcium concentrations observed in these samples were <0.56 mmol/L. In clinical practice, it is unlikely that calcium levels would decrease to such an extent in the majority of trauma or surgery scenarios; however, this is not impossible considering a case requiring significant fluid administration and/or blood transfusion, where intravascular volume expansion or resuscitation is performed with packed red cells, fresh frozen plasma, and saline-based IV fluids. The combination of a large dose of citrate from the blood products and saline intravascular volume resuscitation may lead to an acute hypocalcemia. Such a setting would require careful attention to calcium levels, and these should be treated accordingly.
In some studies, BAL salt solutions have been associated with less coagulation derangement and reduced blood loss compared with saline-based fluids (8–10,19); however, no mechanisms have been reported for these putative differences. Speculation has focused on the role of calcium content in the BAL salt solutions, but no studies have investigated this theory.
Coagulation is a complex enzyme system of activation, inhibition, and lysis of thrombosis involving circulating and tissue-bound coagulation proteases, cofactors, and cellular elements (20). All such enzyme systems require an optimal pH and electrolyte milieu to function optimally. In the coagulation system, calcium remains an important cofactor, but many other pH and electrolyte disturbances could affect the optimal function of this enzyme system. The roles of hyperchloremia and acidosis must be considered in this context. Large volume saline-based IV fluid administration causes hyperchloremic metabolic acidosis (21–23), which in turn has been implicated in coagulation derangements (9,24), increased bleeding (8), gastrointestinal disturbances (11,25), reduced urine output (11), and neurological disturbances (26).
Several limitations of this study need to be considered. First, because it was an in vitro study, it may be of limited relevance to the clinical in vivo setting of fluid resuscitation. The strength of an in vitro study is that it is conducted in well-controlled laboratory conditions, thereby displaying reproducible results. In this study, we were able to clearly examine a dilution range from a clinically relevant dilution to an extreme dilution, enabling the investigation of the potential for coagulation disturbance associated with the various IV fluids. Such a study conducted in vivo would be practically impossible with the number of fluids we have investigated at these dilutions because of the extremely large sample size required to obtain any statistically or clinically relevant answers.
Regarding the choice of analysis, the TEG® was chosen because of its real-time global assessment of coagulation, rather than information of specific coagulation pathways. The drawback of specific pathway analysis in the absence of a global assessment may lead to incorrect assumptions on in vivo coagulation. The global nature of the TEG® analysis remains its great strength. It has been used in many studies like this one, and our group has developed a long track record with its use for other such studies. The device's biggest strength is probably also its biggest drawback, in that the global nature of coagulation assessment does not provide much information on specific coagulation pathways or factors. In the clinical arena, where IV fluids are most commonly used, we consider a global assessment of coagulation important in the management of patients.
The higher 60% dilution may appear to have less relevance to clinical practice. We feel that all the fluid dilutions are necessary to explain a coagulation dose-response relationship of these fluids, because large volume intravascular resuscitation is used in many settings (e.g., trauma and intraoperative hemorrhage). The setting of large volume intravascular resuscitation requires a prudent choice of IV fluid, especially in patients refusing blood product transfusions while undergoing large blood loss surgery. It is here where advanced blood conservation techniques, such as intraoperative normovolemic hemodilution and cell salvage, may be used, and coagulation disturbances need to be limited as much as possible. Under these circumstances, the importance of IV fluid choice cannot be understated.
This study is one of the most comprehensive head-to-head ex vivo comparisons of commonly used IV fluids in clinical practice. The smaller MW, lower substituted starches, as well as the BAL salt preparations, generally exhibited minimal or no coagulation derangement.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. Viuff, B. Lauritzen, A. E. Pusateri, S. Andersen, R. Rojkjaer, and P. I. Johansson Effect of haemodilution, acidosis, and hypothermia on the activity of recombinant factor VIIa (NovoSeven(R)) Br. J. Anaesth., September 1, 2008; 101(3): 324 - 331. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Boldt II. The balanced concept of fluid resuscitation Br. J. Anaesth., September 1, 2007; 99(3): 312 - 315. [Full Text] [PDF]
|
|
|
|
 |
|
 R. M. Peckham, M. T. Handrigan, T. B. Bentley, M. J Falabella, A. D. Chrovian, G. L. Stahl, and G. C. Tsokos C5-blocking antibody reduces fluid requirements and improves responsiveness to fluid infusion in hemorrhagic shock managed with hypotensive resuscitation J Appl Physiol, February 1, 2007; 102(2): 673 - 680. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
