Session III - Combined Meeting with AAST

The Coagulative Response to Reamed Intramedullary Nailing

C.M. Robinson, FRCS; C.A. Ludlam; D.C. Ray; D.G. Swann; J. Christie, Royal Infirmary of Edinburgh, Edinburgh, U.K.

Purpose: To assess whether the extent of activation of the coagulative and fibrinolytic systems correlates with the severity of intra-operative embolic response or risk of postoperative respiratory compromise during reamed intramedullary nailing of tibial and femoral fractures.

Methods: Fat and marrow contents are liberated into the venous circulation during reamed intramedullary procedures. The embolic material is normally filtered from the systemic circulation by the pulmonary microvasculature, but it is thrombogenic and therefore may be involved in the pathogenesis of the post-traumatic "shock-lung" syndromes. To assess the activation of the coagulative and fibrinolytic pathways, we have measured coagulation markers in a prospective series of (Group I, 84 patients) or femoral fractures (Group II, 27 patients) and pathological fractures of the femur (Group III, 23 patients). All patients were also continually monitored using transesophageal echocardiography to assess the extent of embolic events.

Results: The procedures were associated with a significant increase in prothrombin time, activated partial thromboplastin time, prothrombin fragments F₁₊₂ and D-dimers and decrease in fibrinogen, suggesting activation of both the coagulation and fibrinolytic pathways. Evidence of both platelet hyperreactivity and depletion was seen, as estimated by an increase in _-thromboglobulins and a decrease in platelet count. The changes in coagulation markers correlated with the intra-operative embolic response and fall in arterial oxygen partial pressures and saturation (Spearman rank correlation coefficient, P <0.05). Greater changes in these parameters were observed during pathological fracture stabilisation and among patients in whom surgery had been delayed beyond 48 hours (Mann-Whitney U-test, P <0.05). Although operative delay and the presence of intra-operative paradoxical embolization were predictive of the development of postoperative respiratory complications on univariate logistic regression analysis, in a multivariate model that included all variables, only echogenic score and coagulative score were significant independent predictors of severe clinical response (P <0.05).

Discussion: In this study we have demonstrated that reamed intramedullary nailing procedures for fractures are associated with an incremental activation of coagulation pathways. The changes in the measured parameters are consistent with activation of both the coagulative and fibrinolytic pathways with consequent consumption of clotting factors and inhibitors.

An increase in _-thromboglobulins and a fall in platelet count reflected both activation and consumption of platelets. The degree of intra-operative coagulative activation during nailing correlated significantly with the severity of the intra-operative embolic response and fall in arterial oxygen partial pressures and saturation. The greater activation of coagulation associated with high-grade intra-operative embolism could be anticipated given the known coagulative effects of intravasation of marrow contents. The reason for the correlation with the severity of the transient intra-operative fall in oxygen partial pressures and saturation is less clear. It has been suggested that mechanical obstruction of the pulmonary microcirculation by embolic material may interfere with oxygenation or alternatively may produce transient pulmonary intravascular coagulation or vasoconstriction, which could be implicated in this phenomenon.

The severity of both the embolic response and the coagulative profile was greater among patients in whom surgery was delayed beyond 48 hours. Although the deleterious effect of surgical delay has been described clinically among patients with femoral fractures undergoing nailing, quantification of the differences in embolization and activation of the coagulation pathways has not previously been undertaken. The association of disseminated malignancy with coagulation abnormalities is also well recognised, particularly in metastatic adenocarcinoma, where a thrombotic diathesis may be a presenting feature. It is conceivable that the embolic material liberated during reaming is more thrombogenic due to the presence of tumor cells, producing a correspondingly greater activation of the coagulation cascades.

Severe clinical responses were seen in 8 cases. Although severe clinical responses were associated with paradoxical embolization and operative delay, only scores of embolic and coagulative response remained significant when included in a multiple backward stepwise logistic regression model. The effect was additive, and a severe embolic response alone was not associated with clinical effects in the absence of significant activation of coagulation pathways.

Five of our patients with severe clinical responses had a hematological picture consistent with the diagnosis of disseminated intravascular coagulation, and significant changes were apparent in the other 3. Some authors have entirely attributed the respiratory insufficiency to disseminated intravascular coagulation within the lung, producing microthrombosis and pneumocyte damage. Most series report that adult respiratory distress syndrome and disseminated intravascular coagulation often co-exist, with 20% of patients sharing features of both. A generalised coagulopathic state may also explain how the emboli, which are normally filtered from the systemic circulation by the lung parenchyma, may produce signs of systemic embolization in the absence of a patent foramen ovale. As yet however, we cannot establish whether the coagulative changes are primarily involved in the pathogenesis of the severe postoperative responses or are simply a reflection of a more generalised systemic inflammatory response syndrome to the trauma. Several inflammatory mediators are also known to have procoagulant effects and are implicated in the pathogenesis of post-traumatic adult respiratory distress syndrome. The interactions are clearly complex and require further study.

Conclusion: The ability to predict those patients at risk of a severe response following reamed intramedullary procedures is a desirable goal to achieve. At present we are unable to achieve this before surgery with the hematological investigations at our disposal. Further work is required to attempt to identify pre-operatively those factors predictive of the development of severe responses to reamed nailing, so that alternative methods of stabilization could be considered. The differences in response among patients with multiple injuries should also be evaluated. The patients who are at risk of developing a severe postoperative response can be identified by monitoring of the coagulative and embolic responses during surgery. Routine intra-operative transesophageal echocardiography and the assessment of baseline levels of coagulation markers are now used routinely in our unit in such "high-risk" cases, including patients with pathological fractures, patients undergoing delayed surgery and those with known probe patency of the foramen ovale.