Citrated Blood Does Not Reliably Reflect Fresh Whole Blood Coagulability in Trials of In Vitro Hemodilution

Abstract

Hemodilution may have wide-ranging effects on coagulation, and these have been investigated in numerous trials (1–5). Perioperative fluid therapy may leave patients either hyper- or hypocoagulable, with clinical consequences including increased blood loss during major surgery or, conversely, increased risk of coagulation disorders such as deep vein thrombosis. Laboratory coagulation screens are plasma tests based on calcium “flooding” of citrated samples (6). Specific coagulation pathways or factor concentrations can then be assessed. This is acceptable for specific coagulation factor analysis but may not be as appropriate for certain types of whole blood coagulation analysis. Thromboelastography (TEG®; Haemoscope Corp., Skokie, IL) is a point-of-care test of whole-blood coagulation based on the detection of motion of a pin suspended in a rotating cup containing a blood sample and reflecting the coagulation system interactions (7). TEG® is particularly reliable at diagnosing a hypercoagulable state (4,8). This trial forms part of a series of investigations into the coagulation effects of hydroxyethyl starch solutions. Preclinical and clinical studies have suggested an altered coagulation profile of balanced electrolyte and saline-based preparations as judged by TEG®(9). Both increased and impaired coagulation have been shown in the studies, depending on the fluid type used and the degree of hemodilution. Our hypothesis was that recalcifying citrated blood in the TEG® cuvette during trials of in vitro hemodilution would produce results similar to those obtained with fresh whole blood for hemodilution, but that refrigeration of citrated samples may produce accelerated traces. Storage of blood in a citrated medium is widely used in coagulation studies because it reduces the number of venepunctures required during a single day. We therefore performed the hemodilution trial with fresh whole blood, refrigerated citrated blood (4°C), or citrated blood stored at room temperature (21°C). Methods The University of Cape Town ethics research committee approved this study protocol before commencement of the study, and informed consent was obtained from all volunteers. Five healthy volunteers donated fresh whole blood with use of a two-syringe technique from free-flowing veins. The first syringe of blood was discarded. In addition to the fresh whole blood used, a 4.5-mL blood sample was also stored in silicon-coated glass tubes containing 0.5 mL of buffered sodium citrate 0.105 M (3.2%). The fresh whole blood (noncitrated sample) was immediately mixed with a warmed high-molecular-weight hetastarch solution (670 kd/0.75 Hextend® (HEX); Abbott Laboratories, Chicago, IL) at dilutions of 20%, 40%, and 60% in polypropylene tubes (total of 1 mL each) and gently inverted eight times. Undiluted fresh whole-blood control samples were treated in the same way. All samples were placed in TEG® analyzers at 4 min from venepuncture and were allowed to run for 60 min or until the maximum amplitude (MA) was reached. The citrated sample was stored in the refrigerator at 4°C for 90 min, then rewarmed to 37°C. Two hours after the initial venepuncture, the rewarmed, citrated blood was also diluted by 20%, 40%, and 60% with HEX. The samples were treated in the same way as above, except for commencement of the TEG®, where mixed samples were recalcified in each TEG® cup with 10 μL of CaCl2 6.45% immediately after the hemodilution, 1 min before TEG® measurements were started. This volume and concentration of calcium were based on previous preliminary work performed in the University of Cape Town anesthesia laboratory (unpublished data). Limitations on the number of available TEG® channels available required that the volunteers return on a separate day to donate another blood sample. Identical sampling and citrate storage were used. A control sample was again used to ensure a comparable baseline. These citrated samples were left at 21°C for 90 min, rewarmed, mixed with HEX, and analyzed by TEG® by using the same methods as the refrigerated citrated samples previously. Statistics were performed with two-tailed paired Student’s t-tests for data across groups with and without citration. Analysis of variance for repeated mea-sures, with post hoc least significant difference analysis, where significant, was used for multiple testing of the effects of dilution within each group (storage method). Results were considered significant at P < 0.05. Results Citration and recalcification resulted in shortening of the r- and k-times and enhancement of the α-angle after both refrigerated and room temperature storage when compared with fresh whole blood (P < 0.05;Figs. 1–3). The r-time reflects the rate of clot initiation, largely thrombin generation, and the k-time and α-angle reflect the rate of fibrin polymerization. The MA was unaffected. This value is a measure of final clot strength. Refrigerated citrated samples displayed significantly enhanced α-angles compared with nonrefrigerated citrated samples (P < 0.05;Fig. 3).Figure 1: Changes in r-time at various dilutions. Citrated refrigerated samples were stored at 4°C. Citrated room-temperature samples were stored at 21°C. *Significant differences from fresh whole blood at the same dilution; #significant shortening of r-time between diluted fresh whole blood and fresh whole-blood control.Figure 2: Changes in k-time at various dilutions. Citrated refrigerated samples were stored at 4°C. Citrated room-temperature samples were stored at 21°C. *Significant differences from fresh whole blood at the same dilution; #significant shortening of k-time between diluted fresh whole blood and fresh whole-blood control.Figure 3: Changes in α-angle at various dilutions. Citrated refrigerated samples were stored at 4°C. Citrated room-temperature samples were stored at 21°C. *Significant differences from fresh whole blood at the same dilution; #significant increase of α-angle between diluted fresh whole blood and fresh whole-blood control.Refrigerated citrated blood at 20% dilution displayed shorter r-times, along with increased α-angles, than the corresponding fresh blood dilution (P < 0.05). Dilution of 40% revealed shorter r-times and enhanced α-angles in the same groups (P < 0.05), whereas only the r- and k-times were different at 60% dilution (P < 0.01). Citrated room temperature samples versus the fresh blood samples reflected shorter r-times, as well as increased α-angles at 20% and 40% dilutions (P ≤ 0.04), but no differences were observed at 60% dilution. Refrigerated versus room temperature stored citrated samples displayed increased α-angles in the refrigerated undiluted control samples (P < 0.05), and shorter r-times were seen at 60% dilution. Figures 1–3 display these pictures. The fresh whole-blood samples showed a characteristic significant shortening of the r- and k-times, as well as enhancement of the α-angle, at both 20% and 40% dilutions, compared with the undiluted control value, in line with our previous published observations on hemodilution (4,5). Fresh whole-blood MA was reduced at all dilutions. There was no such initial enhancement of coagulation in either group of citrated samples. Compared with its own undiluted refrigerated controls, the refrigerated citrated group exhibited a significantly prolonged r-time, without a significant decrease at any other dilution, at 60% hemodilution (P < 0.05). The k-times were prolonged at 40% and 60% hemodilution (P < 0.05), without shortening at less dilution. The α-angle decrease was significant at 40%–60% dilution, as was the reduction in MA at 20%–60% dilution (P < 0.05;Figs. 1–4). Citrated blood at room temperature displayed prolonged k-times and smaller α-angles only at 60% dilution (P < 0.01), as well as a progressive reduction in MA from 20% to 60% dilution (P < 0.05) versus its citrated room temperature controls (Figs. 2–4).Figure 4: Changes in maximum amplitude (MA) at various dilutions. MA was significantly reduced from the individual controls at all dilutions in all groups. There were no between-group differences.Discussion This study demonstrated that storage of blood samples by using sodium citrate as an anticoagulant resulted in an increased rate of clot formation, particularly if the specimens were also refrigerated. It also showed that the enhancement of coagulation seen after moderate hemodilution cannot be demonstrated in citrated samples. Exploration of the effects of citrated whole blood on trials of hemodilution was necessary, because hemodilution trials with fresh whole blood may require multiple venepunctures of each volunteer during a day. We have previously shown that the method of recalcification used in this study produced accurate and reproducible TEG® measurements of citrated samples compared with control, noncitrated samples, using samples that were recalcified without delay (James MFM, Neil G, unpublished data). On that occasion, we did not investigate the effect of citration on changes due to hemodilution, because we were concerned only with establishing the optimal method of recalcification. In our trial, the pattern displayed by citrated blood was different from that seen with fresh whole blood except for MA measurements. Citration and storage of whole blood resulted in accelerated coagulation that was further accelerated if the stored samples were also refrigerated. Fresh whole blood displays a characteristic response to hemodilution, in which both r- and k-times first shorten at mild to moderate hemodilution (20%–40%) and then increase at 60% hemodilution. The α-angle correctly shows the inverse pattern. The shortened r- and k-times, as well as enhanced α-angles, indicate a faster initiation and rate of clot formation at mild to moderated hemodilution, which is reversed with severe hemodilution (4). In both types of citrate storage, this pattern is lost. This enhanced coagulability with hemodilution is probably “overwhelmed” by the relative hypercoagulability observed with citrate storage of blood, thereby canceling out the subtle effects expected of hemodilution. It is also possible that hemodilution exerts its major effects on the early parts of the pathway that are activated during citrated storage. Camenzind et al. (10) also showed an accelerated TEG® trace in citrated samples versus fresh whole blood stored for longer than an hour; it reached a plateau effect at two hours. The results of Bowbrick et al. (11) were at variance with those of Camenzind et al. (10), reflecting no difference between measured TEG® variables after citration (except k-times and α-angles at 120 minutes in citrated blood stored at room temperature), but theirs was a smaller study with an unquoted power analysis. Interestingly, there was no difference from control even in their refrigerated citrated samples. Refrigeration of whole blood leads to cold activation of both platelets and factor VII (12). It would thus follow that an activated trace would be found with refrigeration. Why the results of Bowbrick et al. (11) did not reflect this is unknown. We observed a significantly faster onset (r-time) and rate of clot formation (k-time and α-angles) of the refrigerated versus nonrefrigerated citrated blood samples. Both citration methods displayed faster onset and rates of clot formation than the fresh whole-blood samples. Sodium citrate acts by chelating calcium, thereby preventing whole blood from clotting. Calcium comes into play only at activation of factor IX (13). This itself is a fair way down the intrinsic pathway. Calcium chelation does not, therefore, prevent contact activation of samples, but rather inhibits activation of the final common pathway. If a citrated whole-blood sample is stored for an hour or two, the upper intrinsic pathway still continues to cascade down to the blocked step of factor IX activation by a small degree of contact activation. This is, in effect, like having a cocked trigger on a revolver. As soon as the sample is recalcified, factor IX is immediately activated, and the rest of the pathway flows on from there. This is why an accelerated trace is seen on the TEG® with citrated samples. Possible limitations of this study are related first to the sample size of five volunteers, second to the return of volunteers a separate day for the final samples, and third to the question of validation of thromboelastography. Variability in results was small, powering the study adequately to detect the differences between citrated and fresh whole blood. The limitations of the TEG® channels available and the running time of each sample necessitated bringing volunteers back a separate day for the final citration method. This could possibly affect data, because blood coagulation could well change from one day to the next. However, our volunteers had control samples taken each day to assess this effect—no significant differences were found. The Thrombelastograph® has been accepted as a test that provides clinically relevant information, and it has been extensively used in trials of hemodilution (both in vitro and in vivo), diagnosing hypocoagulability and, just as importantly, hypercoagulability associated with fluid volume therapy or hemodilution (2–5). Citrated blood is inappropriate for use in trials of in vitro hemodilution. It does not reflect the biphasic dose-response pattern of a hetastarch colloid hemodilution with fresh whole blood. The correlation for the MA (maximum clot strength), though, remains good.