Patient characteristics
One hundred four patients were treated with Interventional Radiology (IR) by endovascular management for PAs over the six-year study period. Thirty-five patients fulfilled the inclusion/exclusion criteria, with the majority excluded for PAs of non-traumatic etiology (e.g. pancreatitis, malignancy, or iatrogenic). Follow-up data after trauma hospitalization ranged from 0 days-4.4 yrs. (median: 44 days; IQR: 2.3–205.5 days).
All 35 patients (10F/25M), average age (± stdev) 41.7 ± 20.1 years, presenting with blunt (n = 31) or penetrating (n = 4) trauma, underwent endovascular treatment with IR. Time from trauma to IR intervention ranged from 2 h - 75 days (median: 4.4 h, IQR: 3.5–17.1 h) with 27 (77%) of PAs identified and treated within 24 h of trauma. Average hospitalization was 13.78 ± 13.4 days. Average ICU stay was 9.2 ± 10.11 days. Ten patients underwent a surgical procedure prior to IR intervention. Seven of the surgical treatments were performed near the site of PA location. Four patients underwent abdominal exploratory laparotomy, followed by treatment of liver or gastroepiploic PAs. Three patients underwent pelvic exploration or orthopedic pelvic/femur fracture repair and were subsequently treated for iliac or femoral PAs. Four of seven patients underwent surgical and IR intervention the same day, ranging from 58 min (direct transfer from the OR) to 7.3 h.
Pseudoaneurysm characteristics
PA number per patient ranged from 1 to 5 (multiple diffuse), with the largest measuring 4.3 × 3.1 cm. PAs were located on the splenic artery (n = 12, 34.3%), pelvic artery (including iliac, femoral, gluteal and pudendal vessels) (n = 11, 31.4%), hepatic (n = 9, 25.7%), upper extremity/axilla (n = 2, 5.7%), and renal arteries (n = 1, 2.9%) (Fig. 1).
Procedural outcomes
Technical success was 30/35 (85.7%). Two patients were not treated based on angiographic and clinical data after discussion between IR and trauma surgeons. Three patients were not treated due to location of the PA. One case was a pediatric patient with an axillary wide-neck PA preventing coil embolization or stent placement, the other two were unable to be treated due to an inability to access the PA or treat safely from a distal location without risk of non-target embolization. Of these five patients, four underwent follow up surgical intervention with treatment of the traumatic region and PA. The remaining 30 patients were treated with coils (n = 6) (Azur Coils: Terumo, Shibya, Tokyo, Japan; Tornado Coils: Cook Medical, Bloomington, IN, USA; Nester Coils: Cook Medical, Bloomington, IN, USA; Concerto Coils: Medtronic, Minneapolis, MN, USA), gelfoam slurry (n = 19) (Pfizer, New York, NY, USA) or both (n = 5). Three patients (8.6%) had PAs with associated arteriovenous or portal-venous fistulas (Fig. 2). All were successfully treated with coil embolization.
Clinical success was 25/35 (71.4%). In patients who were treated, success was 25/30 (82.9%). In all 5 cases of clinical failure, repeat interventions occurred within 3 days of initial treatment and within the same hospitalization (Fig. 3). One patient with an iliac PA treated initially within 24 h of trauma with gelfoam only underwent repeat embolization with gelfoam due to persistent bleeding. The remaining 4 clinical failures required surgical intervention. Three patients underwent splenectomy for persistent splenic PAs. Two of the three patients were treated with gelfoam only while the third underwent embolization with gelfoam and coils. The 4th patient had a gastroepiploic PA treated with Gelfoam as well. That patient required re-exploration due to persistent bleeding with hematoma evacuation.
There was no PA rebleeding or reintervention for PA treatment for any patient after discharge over the reported follow-up periods. Complications occurred in two patients (5.7%). The first complication was classified as an SIR mild AE, consisting of a common femoral PA that resolved without therapy. The second complication was classified as an SIR moderate AE, consisting of pedal punctate arterial emboli that resolved with a heparin drip (moderate escalation of care), without sequelae. Fifteen patients (42.8%) required red blood cell transfusions post-embolization, all occurred within 5 days of IR-intervention (Fig. 3).
Three patients died during the trauma hospitalization for reasons unrelated to the PA. Causes of death included (1) cerebral herniation leading to cardiac arrest, (2) abdominal compression syndrome followed by surgical exploratory laparotomy and profuse coagulopathy bleeding from abdominal wall collaterals, and (3) hypoxemia followed by palliative extubation per the patient’s family’s wishes. All three patients had technically and clinically successful treatment of their PAs.