The mean age was 71.3 ± 14.37 years, and the mean pre-procedural hemoglobin was 7.45 ± 1.62 g/dL, INR pre-treatment was 1.28 ± 1.39. The mean pre-treatment NLR and PLR was 13.9 ± 21 and 188.6 ± 157.4 respectively. The most common cause of LGIB was angiodysplasia

Table 1: Characteristics of our study population and findings on CTA and DSA

Preprocedural computed tomography angiography (CTA) detected active bleeding in 25/32 patients (78,1%).

The angiographic study showed signs of active bleeding in 29 patients (90,6%); in 3 patients (9,4%), no angiographic signs of active bleeding were found.

Table 2: The outcome of PTAE

Super selective embolization was performed in 24/32 (75%) cases, while super selective embolization was performed in 8 cases (25%). TAE was performed in one arterial territory in 20 (62,5%) cases and two arterial territories in 12 (37,5%) cases. Overall, 44 arteries were embolized, corresponding to a mean of 1,4 per patient. The most embolized arteries were the right colic artery (n = 9), the the Ileocolic artery (n = 7) and the cecal artery (n=7).

Fig 1: Types and number of embolized arteries

The most common embolic agent used were a combination of coils and gelfoam (31,3%), a combination of gelfoam and coils (31,3%) and coils (28 %). The mean procedure duration was 67 ± 39 minutes.

In two cases, diagnostic angiography did not demonstrate direct signs of active bleeding, so empirical embolization TAE was attempted. One was treated with super-selective, the latter with selective embolization.

Clinical success was achieved in the super selective - treated - patient, while it was necessary to re-treat the selective one with embolization.

Technical success was achieved in 31 out of 32 patients (96,8%). The only complication encountered in the remaining case was the rupture of an AVM. Clinical success was achieved in 24 patients (75 %) without further interventions (Table 2). In 7 patients (87,5%), recurrent bleeding occurred within a thirty-day window: all of them required required a second embolization (two 24 hours later, one 28 hours later, one 5 days later, one 6 days later, two 14 days later). The proportion of subjects who underwent target or empirical embolization did not differ by rebleeding rate.

Sex, age (< or ≥ 70 years old), coagulopathy (INR ≤ or > 1.5), hemoglobin (< or ≥ 8 g/dL), NLR, active bleeding at MDCT and DSA, bleeding site (SMA or IMA), time between MDCT and DSA were not associated with recurrent bleeding(P > 0.05).

The complication rate (≤ 30 days) was 6,2%; intestinal ischemia, classified as a specific-procedure complication, occurred in 2/32 patients, of whom 1 underwent emergency surgery, with a hemicolectomy performed. It was not possible to treat a patient with sigmoid perforation because he deceased before surgery.

Fig 2: A 61-year-old man presented with severe hematochezia. Computer tomography angiography was positive, showing contrast extravasation from the ileocolic artery approximately 50 cm from the ileocecal valve, and 10 cm distal to the loops treated with previous embolization in the ileocolic territory, (white arrow,A) New shedding and arterial endoluminal contrast agents are recognized in the delayed phases. (white arrow, B) Immediate angiography took place; with microcatheter and micro guide the target vessel was reached and super selective embolization of 2 marginal branches with spirals was performed (black arrows, C-D). The patient underwent right hemicolectomy with ileocolic anastomosis one week after, given the clinical failure of the procedure.

A chi-square test of independence showed no significant association between selective or super-selective embolization and ischemic rate.