HH significantly impairs whole blood coagulation and platelet function in a dose dependent fashion in vitro by reducing platelet function as well as fibrin polymerization. The mechanism can be attributed to the hypertonic saline component and is associated with a dehydration and activation of platelets leading to accumulation of thrombocytes as demonstrated by scanning electron microscopy.
HH is suggested for first line treatment in hemorrhagic shock. Since studies in trauma patients are always affected by an inhomogeneous cohort of patients we have chosen a model of in vitro dilution for standardization of study conditions to estimate the effects of HyperHaes and to identify a possible coagulation impairing substance. Since our study was not designed to evaluate effects on circulatory conditions, we did not adapt dilution volumes of the different agents to possible hemodynamic potentials but in a fixed manner as compared to HH infusion alone. Furthermore the study cannot assess or predict effects on blood loss or outcome.
In vitro studies on coagulation are limited because complex hemostasis pathways cannot be simulated in a complete natural way. Interaction between primary and secondary hemostasis cannot be displayed in coagulation tests. Regular laboratory tests on coagulation use plasma as matrix for analysis. Therefore we decided to use rotational thrombelastometry and multiple platelet aggregation which assay whole blood as a more physiologically matrix to assess coagulation including platelet function. Furthermore thrombelastometry analyzes the end product of coagulation: the clot itself and its stability over time, which indicates clot building potential at the time of analysis. A dynamic time course of coagulation impairment and possible recovery from impairment cannot displayed in our study.
In vivo osmolarity is influenced by numerous factors. Osmolarity in dogs after a 50% blood volume withdrawal and following infusion of 4 ml*kg-1 hypertonic NaCl (2400 mOsmol*l-1, which is comparable to HyperHaes) led to an increase of plasma osmolarity from 307 mOsmol*l-1 to 333 mOsmol*l-1 within 30 minutes . Estimating average plasma osmolarity of 300 mOsmol*l-1 and an osmolarity of 2400 mOsmol*l-1 for HyperHaes in vitro dilution by 5% would suggest a resulting osmolarity of approximately 405 mOsmol*l-1 which is already markedly above physiological levels. These in vitro high osmolarity conditions could compromise the translation of the results into clinical settings. Nevertheless, it remains unclear if compensation mechanisms are able to adjust osmolarity before interfering with platelets. In a different setting of acidosis and diminished coagulation laboratory parameters did not return to normal after compensation of acidosis . Furthermore it could be possible that repeated administration or overdosage of HH could account for a non-physiological increase in osmolarity exceeding possibilities of compensation.
Normal blood volume in adults may be estimated to be 70 - 80 ml/kg bodyweight. Accordingly, the recommended HH dose of 4 ml/kg bodyweight in patients with hemorrhagic shock yields a hemodilution of 1:17.5 (5.7%) to 1:20 (5%). Since this mirrors normal conditions without blood loss we have chosen a 5% dilution as lowest degree of dilution for our study. Blood loss would lead to a further reduction in circulating blood volume and thus to a relatively increased portion of infused HH per ml blood volume resulting in an increased test agent/blood ratio, ergo to greater dilution. Blood loss of 50% blood volume then would lead to approximately 1:10 (10%) dilution, 75% blood loss would account for a 1:5 (20%) dilution and 40% dilution would be comparable to 87.5% blood loss. With respect to this consideration increasing blood loss would lead to increasing relative overdosage accounting for possible enhancement of otherwise induced coagulation disorders.
Even 5% whole blood dilution with HH significantly impaired platelet function. This effect on thrombocytes cannot be adequately detected in whole blood coagulation. However, MCF was affected in all samples with ≥10% dilution and CT prolongation finally occurred when dilution was 40%. Maximum clot firmness in whole blood coagulation is basically determined by platelet function and fibrin polymerization, while clotting times are dependent on the speed of thrombin generation by clotting factors . Thus, HH affected platelet function and fibrin polymerization in a more severe way than action of clotting factors. Responsible for interference with fibrin polymerization of HH is its HS portion, since we demonstrate a comparable impairment of fibrin clot firmness by HH as compared to HS. It is well known that HS inhibits fibrin polymerization [14–18]. Our data are consistent with these findings. This effect is most likely caused by dilution of fibrinogen  and decreased FXIIIa-mediated fibrin cross linking [14, 15]. However, the precise molecular mechanism still remains unclear.
The mechanism of action of HH to improve blood pressure is based on mobilization of extravasal fluids along an osmotic gradient by intavasal administration of HH . We suspected this intavasal hyperosmolarity also to be one possible mechanism of interaction between the hypertonic solution and platelets leading to dehydrated and functionless thrombocytes. Platelets treated with and without HH were examined by electron microscopy. In the HH dilution deformed single platelets as well as large aggregates of activated platelets can be seen (figure 2). Such aggregates could account for a loss of platelet function and in vivo could lead to an obstruction of small vessels leading to a reduced platelet count as well. A detection of such aggregates after in vivo administration of hypertonic saline solution has not been done to date and would be of great interest concerning our findings.
In experimental settings controversial effects of HH on coagulation have been described. In animal models of uncontrolled hemorrhage treatment with hypertonic saline led to an aggravation of hemorrhage [21–23]. In these studies only hypertonic saline was studied while HS was not administered alone or in combination with hypertonic saline. In a recent study in a model of uncontrolled hemorrhage in pigs after liver injury less hemorrhage after HH administration was observed as compared to the use of colloids alone . However, in this study red blood cells collected by an automated cell saver were simultaneously to the test agent infused. As a consequence the dosage of the hypertonic and hyperoncotic agent was reduced in a relative way by the parallel infusion of red blood cells which could have weakened the coagulation impairing effect of HH. Despite this, to reflect comparable hemodynamic potential greater volumes of colloid infusions were admitted leading to a higher dilution of clotting factors in the control group. Since red blood cell concentrates or cell saver blood is available in the hospital only the settings of this study are more comparable to an admission in the emergency room or the operating theatre than to a preclinical situation. As a consequence conclusions on the influence of these solutions on coagulation and blood loss in a preclinical situation should be drawn with caution.
Another hazard might occur when hypertonic saline is used in combination with large doses of colloids due to additional risks of adverse effects of colloids itself as for example anaphylactic reactions or reduction of kidney function which also have to be considered [24–27].
In different clinical situations of major blood loss such as penetrating chest trauma , patients undergoing cardiac surgery [29, 30], or vascular surgery [31–33] studies indicating beneficial effects on outcome have been published. However, results of meta-analysises showed if any only minor improvement of survival no matter if hypertonic saline solution is used exclusively or in combination with colloids [34–36].
Our results indicate HH to cause a dose dependent impairment of platelet function and whole blood coagulation. However, these effects appear to be small in dilutions comparable to expected dilution after treatment of shock when the circulating blood volume is not reduced. From a different point of view this implicates that considering a small therapeutic index the risk of overdosage seems to be high and should be strictly avoided. Whether this also accounts for repeated admission and length of a time interval for possible safe repeated administration of HH cannot be assessed in the present study and may be addressed in future investigations.
Furthermore, the recommended dosage of HH is calculated with respect to bodyweight. In clinical situations variables as for example body weight can be assessed easily. In preclinical situations it is much more difficult to assess the patient's bodyweight which could lead to overdosage per se.
We calculated our dilution series to compare resulting dilution effects to HH treatment at different degrees of severe blood loss. Since we found greater effects on platelets with increasing dilution due to higher drug levels, we suspect HH treatment to show increasing negative effects on coagulation and platelet function with increasing blood loss due to possible relative overdosage. HH is designed to help stabilizing circulatory conditions in these situations. This implicates that dosage in patients with higher blood loss should be calculated with care, repeated administration should be avoided and the physician should be aware of increasing coagulopathy.
Since it remains questionable if our findings can be transferred into clinical settings clinical studies are necessary to evaluate such issues.