Excessive perioperative bleeding: are fibrin monomers and factor XIII the missing link?

PREDICTING perioperative blood loss is traditionally fraught with great difficulties. Medical history is certainly crucial in the risk assessment of perioperative blood loss because it provides information on bleeding history, the use of antiplatelet drugs, and oral anticoagulation. However, so far there is no laboratory parameter that could reliably predict perioperative blood loss. According to the paper of Korte et al. in this issue of ANESTHESIOLOGY, there may be hope at least for surgical cancer patients. In these patients, increased fibrin monomer levels, suggestive of preoperative coagulation activation, might be predictive of increased intraoperative blood loss. The hypothesis of the authors is that the increase in fibrin monomer would be due to a preexisting imbalance between available factor XIII and thrombin. Thrombin proteolytically cleaves fibrinogen to soluble fibrin monomers, which are incorporated into a crosslinked and thus stabilized or nonsoluble fibrin network in the presence of appropriate amounts of activated factor XIII. Thrombin also cleaves the A subunit of factor XIII to generate the enzymatically active factor XIII (FXIIIa), capable of cross-linking fibrin monomers into a fibrin strand (fig. 1). The authors hypothesize that a relative imbalance of thrombin in respect to activatable factor XIII would result in an accumulation of fibrin monomers due to the inadequate incorporation of monomers into a stable fibrin network. They further hypothesize that in patients with preoperatively elevated fibrin monomers, the surgical procedure may lead to an accentuation of the above-described imbalance that may be associated with increased perioperative blood loss. The authors tested their hypothesis in a pilot study with 22 patients. Intraoperative factor XIII substitution resulted in a reduced decrease of fibrinogen concentration, a smaller decrease of mean clot firmness in rotational thrombelastography analysis, and reduced blood loss. Interestingly, the authors were able to demonstrate in an earlier study that the developing clot had reduced maximum clot firmness before the exaggerated surgical blood loss was observed. Most fascinating, by a single dose of factor XIII (30 U/kg) given 15 min into surgery, the exaggerated decrease of maximum clot firmness during surgery could largely be prevented, and this resulted in a reduced blood loss. A potentially relevant characteristic of the study population of Korte et al. is that patients had pathologically low factor XIII activities at baseline (approximately 58% in the verum group vs. 55% in controls). The study population thus presented with increased fibrin monomer levels and (putatively) acquired factor XIII deficiency. Whether intraoperative factor XIII substitution will prove useful in patient populations without preexisting acquired factor XIII is an open question. A low factor XIII level has been shown, in keeping with the results of Korte et al., to be a risk factor for hemorrhage after intracranial and coronary artery bypass surgery. With factor XIII treatment, chest tube drainage and transfusion needs were indeed reduced in patients with a normal postoperative factor XIII level. The study by Korte et al. is highly interesting and thought-provoking. If the conclusions hold true despite significant methodological shortcomings and findings that are difficult to interpret, we may in the future test patients for fibrin monomers preoperatively, and treat them if elevated with (a single dose of) factor XIII early into surgery. At the moment, such a conclusion is premature because this study has significant limitations. First, only 22 patients were analyzed. This limits the power of the study, and the sample size is far too small to allow any judgment on safety. All procoagulatory substances may be associated with adverse thrombotic and thromboembolic events. Second, the study was powered to show a difference of 30% in maximum clot firmness and was prematurely terminated because this threshold was achieved at a planned interim analysis with a P value of 0.04. Stop rules at planned interim analyses usually use significantly lower P values to avoid a false positive result due to an (initial) uneven patient allocation to the groups by chance and to preserve an overall two-sided type I error rate of 0.05. Therefore, the question arises whether this trial may have been This Editorial View accompanies the following article: Korte WC, Szadkowski C, Gähler A, Gabi K, Kownacki E, Eder M, Degiacomi P, Zoller N, Devay J, Lange J, Schnider T: Factor XIII substitution in surgical cancer patients at high risk for intraoperative bleeding. ANESTHESIOLOGY 2009; 110:239–45.