Fat embolism occurs frequently and can be detected by means of trans-esophageal echocardiography in more than 90% of patients suffering from fractures of the long bones [
1]. On the other hand, the incidence of FES is considerably lower: in a study of 274 consecutive patients with isolated femoral shaft fractures, Pinney
et al.[
4] could show an FES rate of only 4%. Analysis of the subgroups showed development of FES manifestations in all patients below the age of 35 as well as in patients in whom treatment had been initiated more than 10 hours after trauma. Our work also reports on a patient under the age of 35, but surgery commenced within six hours of the trauma. The incidence of FES is considerably increased in patients suffering from multiple injuries [
2]. In a series of 211 patients suffering from multiple injuries, Riska and Myllynen [
5] only found three patients (1.4%) who received surgery; however, one patient died. On the other hand, 84 patients (22%) in the comparison group received conservative treatment. Apart from emerging from fractures [
6,
7], FES can also be caused iatrogenically by intramedullary nailing of the femur or the tibia. It is assumed that fat particles are introduced into the venous system as a result of increased intramedullary pressure caused by the intramedullary pin, which will almost always result in the formation of droplet-shaped fat agglomerations in the capillary areas of the lungs. This formation will generally lead to pulmonary micro-embolism resulting in increased perfusion pressure, congestion of the lung vessels and secondary overstressing of the right side of the heart, which in turn may result in hypoxemia, probably with acute right-sided heart failure. Furthermore, the bone marrow in the venous vessels causes considerable activation of coagulation with a decrease in thrombocytes and consumptive coagulopathy (disseminated intravascular coagulation). Petechiae (punctuate bleeding) may appear on the trunk of the body as well as sub-conjunctivally as a delayed effect. However, this clinical characteristic was not observed in our patient. The maximum pressure measured during the reaming of the medullary cavity in preparation for a femoral intramedullary pin may reach 400-500 mmHg [
8]. These pressure values are primarily achieved during the opening procedure and the first drill sizes. If the medullary cavity is sufficiently widened, the procedure of screwing in the pin will not cause excessive pressures anymore. Screwing the intramedullary pin into an unwidened medullary cavity will lead to pressures of 200-300 mmHg [
9]. Here, the screwing process does not cause any increase in pressure; however, screwing in the pin will lead to pressure values as high as those reached during the drilling process. For the prevention of FES, no significant differences were found with regard to the femur, that is whether intramedullary pins were introduced into a widened or an unwidened medullary cavity [
10]. Paradoxical FES will occur if the origin is initially located in the venous system, and arterial circulation takes place prior to potential pulmonary manifestation. Potential causes for such manifestations are, for example, latent or patent foramen ovale [
11], ventricular septum defects, persistent truncus arteriosus, arteriovenous malformations, or - as in our patient - an atrial septal defect in high position with right-to-left shunting. However, only very few case reports on paradoxical FES are available in the literature. Christie
et al.[
10] reported on four patients with latent foramen ovale, who developed paradoxical FES because of the reaming of the medullary cavity of the femur; two out of these four patients died. The intravasations were documented intra-operatively by means of trans-esophageal echocardiography. Kallina and Probe [
12] reported on a 20-year-old female patient with previous mitral valve prolapse, who developed paradoxical FES after fractures of the femur and the tibia. Reaming of the respective medullary cavity was conducted 16 hours after trauma, prior to intramedullary nailing. In contrast to our patient, a decrease of oxygen saturation was noted on the already awake patient at the end of surgery, leading to intubation. Similar to our patient, diagnostic investigation showed cerebral ischemic disorders with white, matt stipples as well as generalized spasticity. In contrast to our patient, this patient was completely oriented again after 55 post-operative days, and speaking did not present a problem to her. Although embolism was not documented intra-operatively by means of echocardiography in our patient, paradoxical cerebral embolism had to be suspected because of the high-positioned atrial septal defect with right-to-left shunting, which had not been diagnosed before. Pulmonary deterioration was not observed at any time, neither diagnostically nor clinically. Finally, the hypothetical question remains whether FES was caused by the femoral fracture itself or by intramedullary nailing. There is evidence indicating that both femur fractures and intramedullary nailing lead to introduction of fat into the circulatory system, not only on their own but also in combination. In our patient, this combination resulted in fulminant paradoxical FES, therefore the authors recommend plating of femoral fractures instead of nailing.