In a porcine model, we have shown that intraosseous fluid resuscitation causes systemic fat embolization in addition to pulmonary emboli. Our findings suggest that intraosseous cannulation and/or infusion may cause widespread pulmonary fat emboli, supporting current literature [6, 24].
Three control animals also developed pulmonary fat emboli, which we believe came from the median sternotomy, a procedure known to cause fat embolization [25–27].
Intraosseous fluid resuscitation caused fat emboli to be dislodged in the coronary vessels of the left ventricle in animals with closed chests. To our knowledge, systemic fat embolization confirmed by biopsies following intraosseous cannulation and/or infusion has not been described in an animal model. Systemic fat embolization following intraosseous infusion has been documented in one single clinical case [12]. We found cardiac fat emboli only in pigs with closed chest. This may suggest, with the limitation of few pigs in the series, that an open chest does not reduce the threshold for when fat reaches the systemic circulation through the lungs, as is the case concerning air emboli [22].
The pigs did not develop clinical fat embolism syndrome for the duration of the experiments, a finding consistent with previous literature [24]. We observed significant ST-elevations in three animals who were found to have coronary fat emboli. The lack of major clinical deterioration may be explained by the fact that fat embolism syndrome usually occurs 24 to 72 hours after the embolic event [15, 28–30].
We analyzed several anatomical separate slices of each animal’s brain and kidneys but did not find convincing evidence of intravascular fat emboli. The fact that systemic fat emboli were only found in coronary vessels may be random, and the lack of emboli in the brain and renal biopsies does not exclude the possibility that embolization to these tissues occurred. More extensive biopsies could increase the sensitivity in discovering further systemic embolization, and cerebral magnetic resonance imaging (MRI) with susceptibility weighted imaging (SWI) would to a high degree of certainty reveal fat embolization to the brain [14, 31–33].
Systemic embolization
Systemic passage of emboli occurred in several animals, in all cases in the form of coronary embolization, and only in animals with closed chest. Systemic passage – or shunting – occurred in the closed chest group as evident by the presence of coronary fat emboli. None of the included animals had an intracardiac shunt. Systemic passage of emboli in the absence of intracardiac shunt has been shown in humans and animals [19, 34–37]. Venous emboli can theoretically enter the systemic circulation through at least three sites; the pulmonary capillary network, an intracardiac shunt – for example an open foramen ovale - or through extra-alveolar shunt-vessels, more commonly referred to as anatomical shunts [34, 38]. It has been proposed that intravascular fat emboli may be “squeezed through” the pulmonary capillary network, i.e. that small pulmonary fat emboli may migrate through the pulmonary capillary network and cause systemic embolization [19].
In addition, extra-alveolar shunts exist in healthy humans [34, 38, 39], likely contributing to systemic embolization. The extent of shunting through these vessels is probably dynamic, as the shunting appears to increase with hypoxia, posture change, and exercise [34, 38, 39].
In our experiments, we hypothesized that an open chest would increase systemic fat embolization. We believed that an open chest would cause reduced transmural lung pressure, allowing over-expansion of the lungs. This, in conjunction with mechanical ventilation, has been proposed to facilitate the recruitment of extra-alveolar shunt vessels, thus opening up dynamic passages available for systemic shunting for air emboli [22, 40]. Fat emboli likely do not share the gravitational, thrombogenic, or physical properties of air emboli, and are probably influenced by the mentioned factors in other ways.
Clinical relevance
We have demonstrated that intraosseous infusion causes pulmonary and potentially coronary fat embolization, and that systemic embolization did not occur in animals with open chest. Intraosseous infusion is used when intravenous access is difficult, as is often the case in patients with serious hypovolemia or hypotension. In critically ill patients, delayed fluid or drug administration may be fatal. Intraosseous access may be warranted in time-critical situations, despite a high probability of pulmonary emboli and a risk of coronary and cerebral embolization, as these emboli may commonly be subclinical. Thus, open chest situations, as with thoracic trauma and emergency thoracotomy, do not by itself seem to contraindicate intraosseous cannulation.
Limitations
Our study has potential limitations. A subgroup of pigs was exposed to both sternotomy and intraosseous infusion, and both interventions are known to cause fat embolization, possibly placing these pigs at a higher risk of systemic passage of fat emboli. Despite this, no coronary fat emboli were observed in pigs with open chest. Thus, it is reasonable to conclude that the intraosseous infusion, and not the sternotomy, was the most significant source of fat emboli in both pigs with open and closed chest. We conducted the experiments at two laboratories, possibly introducing a bias in our findings. However, as all experiments were conducted by the same researchers using identical protocols, medications, and equipment is is not likely that this has affected our findings to any extent. Further, the animals were of the same race and bred from the same national insemination stock and supplier. Echocardiography was performed by two different researchers, and neither detected passage of systemic fat emboli, even in cases where fat emboli were found in postmortem left ventricle biopsies, suggesting that the TEE had a low sensitivity for the detection of the systemic passage of fat emboli.
The distribution of systemic fat emboli in the tissues is unpredictable, and despite acquiring several tissue biopsies postmortem, we sampled only a fraction of the organs, and organ embolization may indeed have occurred without us detecting them in the sampled tissue. Finally, it is established that clinical fat embolism syndrome usually does not manifest itself until 24–48 hours after fat embolization has occurred [15, 30, 41]; our observational period was limited to 300 minutes. Thus, we cannot exclude the possibility that the fat emboli would later have caused fat embolism syndrome.