Major pelvic fractures result from high-energy trauma, including traffic accidents and falls, which usually leads to multiple injuries complicating the patient’s management.
Pelvic fractures are present in 4 to 9% of patients with blunt trauma and are associated with significant morbidity and mortality. Patients that present in shock have a mortality of greater than 30% in modern series (1).
Management of these patients requires a coordinated approach between trauma surgeons, orthopaedic surgeons and interventional radiologists.
Endovascular interventions can be lifesaving in patients with haemorrhage from pelvic fractures, however, in the literature there is currently no univocal consensus on the role of angiography in the hemodynamically unstable patient. In most studies, preperitoneal pelvic packing with external fixation has been demonstrated to decrease mortality in patients with pelvic fractures presenting in shock; it is reserved only for extremely unstable patients and is not effective for arterial haemorrhage control.
The pelvic vasculature and viscera are in close continuity with the bony pelvis; high-energy pelvic ring disruptions are often associated with significant haemorrhage. The bony anatomy of the pelvis can be conceptualized as a ring formed by the sacrum and right and left innominate bones, specifically the ischium, ilium, and pubis. Paired internal iliac arteries and their related tributaries are the predominant arterial supply of the pelvis. The venous system follows a similar path but is arranged in a plexus adherent to the posterior pelvic wall. Pelvic fracture bleeding predominantly arises from the venous plexus or cancellous bone; however, arterial bleeding occurs in a significant number of cases and is associated with life-threatening haemorrhage. Knowledge of pelvic fracture classification is useful in predicting the likelihood of severe injury and aiding communication with consultants. A variety of classification systems have been proposed (2); the two most commonly described classification systems in emergency medicine literature are the Young-Burgess (YB) and Tile fracture classification systems.
Imaging diagnosis is mostly obtained by a contrast-enhanced CT scan, especially in stable patients or patients who are responsive to resuscitation.
Contrast-enhanced CT is fast and provides detailed information about the type and location of the haemorrhage; the presence of vascular injuries such as pseudoaneurysm, arteriovenous fistula, or complete vascular transection, and the presence of vascular anatomic variants. The development of widely available fast scanners has resulted in a change in clinical practice, and CT scans are currently performed even on hemodynamically unstable patients.
The sensitivity and specificity of CT with contrast for localizing arterial extravasation in pelvic trauma is 60 to 90% and 87 to 98%, respectively (4).
Clinical indications for pelvic angiography in a patient with pelvic traumatic fracture include (1) persistent hypotension after 2 L of fluid challenge with exclusion of intraperitoneal haemorrhage, (2) transfusion requirement exceeding four units of blood within 24 hours, or six units within 48 hours, (3) a pelvic hematoma greater than 600mL or active extravasation seen on CT, and (4) a large or expanding pelvic hematoma found on laparotomy. Furthermore, angiography and transcatheter embolization should be guided by the hematoma volume, rather than by the presence or absence of contrast extravasation alone.
If embolization is indicated, the choice of embolizing agents and exact embolization strategy depend on angiographic data, vascular anatomy and haemodynamic stability.
There is not enough evidence to suggest the use of one embolic material over another, however, embolic material is chosen according to the bleeding site and the presence or absence of coagulation abnormalities. Microcoils seem to get good results, and liquid agents such as Onyx may be more effective in coagulopathic patients. Whenever possible, embolization is performed with temporary material, and Gelfoam pledgets or slurry are usually the agents of choice in multifocal injuries or brisk haemorrhage. In haemodynamically unstable patients and in cases of multiple bilateral bleeding targets, non-selective bilateral embolization of the internal iliac arteries should be performed. Non-selective unilateral internal iliac artery embolization is necessary in case of multiple unilateral bleeding targets or in case of failed attempts at selective embolization. Selective embolization should be performed in hemodynamically stable patients with one or few bleeding targets identified on the CT scan or angiography (3). IR unit organization and timing are, furthermore, crucial to enhance outcome parameters. My current practice provides 24/7 on-call service to perform angiography and embolization within 30 minutes of the activation of a trauma code because a shorter door-to-embolization time is associated with better survival in the first 24 hours.