Reversal of colloid-induced coagulopathy

Unfortunately, there is no drug in our critical care repertoire that has no potential risks or side effects. Risk awareness, co-administering prophylaxis, and targeted symptomatic treatment of side effects are routine strategies. Some medications are combined in one pill, e.g. the active painkiller with an adjuvant substance against the common side effect of obstipation. Such a pharmacological approach appears to be acceptable if efficacy, safety, and galenic compatibility of the active compound and the adjuvant are proven, and if no alternatives without such side effects exist.

If colloids are infused restrictively and according to individualized preload and/or microcirculatory targets (and if colloids are not abused for general fluid substitution), dose-dependent side effects on clot strength are suggested to be minimal and – most likely – not requiring reversal. Nevertheless, it is intriguing to consider prescribing colloids – in order to use their superiority in volume efficacy over crystalloids – together with fibrinogen concentrate in order to reverse eventual colloid-induced reductions in clot strength. Feasibility studies in in vivo animal experiments indicate that fibrinogen concentrate can rapidly reverse MCF to baseline values. In vitro, however, HES-induced FIBTEM reductions could be reversed completely by fibrinogen concentrate or cryoprecipitate in some experiments but not in others.[25,26] HES-induced MCF reductions responded less to fibrinogen concentrate compared with gelatins or albumin (also at hypothermia), and the combination with factor XIII concentrate improved reversal. These experimental findings suggest the use of fibrinogen concentrate after resuscitation with albumin and gelatins.[27] ROTEM parameters cannot be improved in vitro with factor XIII concentrate alone in any tested diluent.[27]

No clinical and histopathological signs for thromboembolic events could be detected, suggesting safety of factor concentrates for reversal.[26] Similarly, thrombin generation potential was not increased by fibrinogen concentrate addition, whereas it was increased for cryoprecipitate and plasma, suggesting thrombogenic risks of a reversal approach based on allogeneic blood products.

Clinical relevance of colloid-induced coagulopathy

The question arises whether disturbances in laboratory parameters translate into clinically relevant bleeding manifestations, burdens, and harm, and whether fluid-induced coagulopathy is an independent risk factor for anemia, blood transfusion, and mortality, as well as a driver for resource use and costs. These aspects of patient safety and healthcare economics have received more attention since the World Health Organization (WHO) started supporting the multimodal therapeutic concept of patient blood management (PBM). Inherent side effects of fluids on dilution-dependent anemia and dilutional coagulopathy may counterbalance the role of fluids in PBM (to increase the tolerance to anemia) and, therefore, need careful consideration. PBM requires clinicians worldwide to apply a restrictive transfusion practise and to use physiological triggers for transfusion of pRBC instead of hemoglobin triggers.

Outcome studies report controversial results. A meta-analysis published in 2013 detected no adverse effects of tetrastarch in the surgical population: in 38 trials (3,280 patients) no increase in blood loss was found, in 20 trials (2,151 patients) no increase in allogeneic blood transfusions was found, and there was no signal for increased mortality.[28] This finding was confirmed by a recent meta-analysis.[29] More recent evidence confirms no increase in blood loss and transfusion requirements after HES exposure in major abdominal surgery [17] and cardiac surgery.[19,21,30] In neurosurgery, blood loss was not increased in patients receiving HES.[15,16] The intraoperative infusion of HES was not associated with the higher incidence of post-craniotomy intracranial hematoma formation requiring surgery, while perioperative arterial hypertension and the use of non-steroidal anti-inflammatory drugs were risk factors.[31]

Some studies, however, demonstrate increased transfusion requirements despite similar blood loss or intact coagulation in cardiac surgery.[18,20] The explanation may be the use of hemoglobin levels as transfusion triggers instead of physiological transfusion triggers; because of the higher volume efficacy, hemoglobin drops more after colloids than after crystalloids. Transfusion requirements, mortality, length of stay, and infections were all observed to be reduced in orthopedic patients resuscitated with HES.[32] However, some authors reported increased blood loss on implementing fixed-dose rather than goal-directed protocols.[33,34] Retrospective analysis suspected increased volume efficacy-related hemo-dilution as a contributor to increased blood loss and transfusion requirements in patients receiving HES versus crystalloids.[35]

In the critically ill and septic population, studies again reported controversial results. Mean increases of 18 ml FFP in HES-treated patients appear not clinically meaningful, despite being statistically significant.[36] In septic patients receiving tetrastarch, absence of an increase towards hypercoagulability in the TEG was hypothesized to be predictive for death and bleeding.[37] Some trials showed no effects of HES on blood loss,[38] while others found an increase in major bleeding events and transfusion rates.[37,39] However, in the latter studies, indication and dosing of the tetrastarch has been criticized; it was not the drug per se that was considered harmful, but rather the way it had been used.[40] Post-hoc analyses and meta-analyses aggravate methodological concerns and hide them behind the claim of high-quality evidence-based medicine. While waiting for refined trials considering denominators of quality of critical care, we have to acknowledge the alarming signs in the existing trials indicating deleterious effects of HES: patient safety management warrants the avoidance of HES in critical illness.

Increased mortality has been reported in those patients developing HES-induced clot strength reductions:[41] reduction in thromboelastographic maximum amplitude was an independent predictor for mortality. This is surprising, because, in general, viscoelastic tests have a poor predictive value and are therefore used and recommended for detecting actual pathomechanisms of bleeding manifestations rather than for prediction purposes.[2] Even more surprising is the finding that minimal changes in absolute values of thromboelastographic clot strength within the reference range have been suggested to be predictive for survival.[39] Future prospective trials will have to confirm this observation or identify potential statistical random failure. A recent meta-analysis found no increase in mortality, incidence of the need for renal replacement therapy, bleeding volumes, or in transfusion requirements in non-septic patients at the intensive care unit [42].