Collateral Vessel Framework Guides Embolization for Inferior Epigastric Artery Bleeding

In an observational study evaluating endovascular treatment for arterial bleeding, researchers at Gunma University Hospital demonstrated that transcatheter arterial embolization achieved universal technical success in occluding target vessels, while clinical recovery diverged sharply depending on vascular collateral anatomy and patient physiology. Investigating inferior epigastric artery hemorrhage across diverse medical contexts, the clinical investigators established that the vessel does not function merely as an isolated conduit within the abdominal wall, but frequently serves as a critical bridge within extensive pelvic circulatory networks. Across a retrospective cohort of sixty-two consecutive patients treated between January 2011 and December 2024, interventional radiologists recorded an overall technical success rate of one hundred percent, confirming complete angiographic disappearance of active extravasation or pseudoaneurysms at the end of each procedure. However, the overall clinical success rate reached 82.3 percent, with outcomes varying markedly across four distinct collateral pathways identified on diagnostic imaging.

The inferior epigastric artery arises from the external iliac artery just superior to the inguinal ligament and ascends within the rectus sheath, where it forms extensive anastomoses with the superior epigastric artery to create a craniocaudal epigastric arcade. Beyond its course in the anterior abdominal wall, the vessel gives rise to caudal pubic branches that frequently connect with the obturator artery along the superior pubic ramus to form an arterial corona mortis, linking the external iliac system to the internal iliac territory. In female anatomy, the artery additionally supplies the round ligament artery, which integrates into a dense collateral vascular network surrounding the uterus and pelvic venous plexus. Because of these complex connections, injuries involving the inferior epigastric artery occur not only as spontaneous or anticoagulation-induced rectus sheath hematomas, but also as severe complications of pelvic trauma, obstetrical delivery, and percutaneous abdominal punctures, complicating standard transcatheter hemostasis.

To address the therapeutic challenges posed by this variable vascular anatomy, the investigators devised a collateral-pathway-oriented clinico-angiographic framework that classifies bleeding into four distinct functional categories based on contrast-enhanced computed tomography, digital subtraction angiography, and clinical context. Type A represents an isolated abdominal wall pattern, characterized by hematoma confined to the rectus sheath or adjacent parietes without pelvic fractures or round ligament involvement. Type B describes a round-ligament or postpartum pattern, featuring active extravasation around the uterus and round ligament in the setting of postpartum hemorrhage, often persisting after initial uterine artery embolization. Type C encompasses a corona mortis or pelvic-fracture pattern, where bleeding originates near pubic rami fractures via anastomotic pubic branches or communication with the obturator system. Type D identifies iatrogenic puncture-related injuries, defined by direct trauma to the arterial trunk or superficial branches following catheter placement, abdominal paracentesis, or drainage procedures.

The study protocol mandated rigorous diagnostic imaging before and during endovascular catheterization to map the responsible arterial architecture and confirm active bleeding. Initial examinations utilized multidetector contrast-enhanced computed tomography under a standardized emergency protocol, administering intravenous iodinated contrast material at concentrations of 300 to 370 milligrams of iodine per milliliter and injection rates of 3 to 4 milliliters per second. Scanners utilized thin collimation of 0.5 to 1.25 millimeters, triggering arterial phase acquisitions using aortic bolus tracking at approximately 150 Hounsfield units, followed by portal venous phase acquisitions delayed sixty to seventy seconds after injection initiation. Transcatheter arterial embolization was subsequently executed under local anesthesia using a common femoral artery approach, advancing diagnostic catheters into the external iliac artery before deploying coaxial microcatheters over guidewires into culprit branches to verify extravasation under selective digital subtraction angiography.

Interventional teams selected specific embolic materials tailored to the underlying clinico-angiographic classification, hemorrhage volume, and baseline systemic coagulation status. Gelatin sponge particles served as the primary embolic agent, utilized alone in forty-eight patients, accounting for 77.4 percent of the entire study cohort, particularly for diffuse oozing or when superselective catheter seating was established. Pushable or detachable metallic coils, deployed either alone or in combination with gelatin sponge, were utilized in eleven cases, or 17.7 percent of procedures, primarily to achieve focal exclusion of discrete pseudoaneurysms or secure distal vascular occlusion. In three patients, comprising 4.8 percent of cases, interventionalists utilized N-butyl-2-cyanoacrylate mixed with ethiodized oil at ratios between 1:2 and 1:4, flushing microcatheters with five percent dextrose solution prior to delivery to prevent premature polymerization and provide rapid, coagulation-independent hemostasis during profound instability or severe coagulopathy.

Evaluation of clinical outcomes across the sixty-two patients demonstrated substantial disparities despite uniform angiographic occlusion. The overall clinical success rate—defined as sustained hemodynamic stabilization within twenty-four hours after embolization without requiring repeat transcatheter embolization or surgical hemostasis—stood at 82.3 percent, representing fifty-one patients. Patients categorized under Type A experienced a clinical success rate of 96.4 percent, with twenty-seven of twenty-eight individuals stabilizing successfully. Similarly, all three patients presenting with Type D iatrogenic puncture injuries achieved clinical success, yielding a rate of one hundred percent. In contrast, intermediate clinical success occurred among Type B postpartum hemorrhage patients, where twelve of fifteen individuals stabilized, representing 80.0 percent. The lowest clinical success occurred among Type C pelvic fracture patients, where only nine of sixteen individuals achieved the primary clinical endpoint, resulting in a success rate of 56.3 percent.

The composite endpoint of clinical failure, which encompassed a lack of clinical success or in-hospital death from any cause during the index hospitalization, occurred in eleven patients overall, representing 17.7 percent of the study population. In-hospital mortality reached 12.9 percent across the cohort, accounting for eight patient deaths. These fatalities were concentrated almost entirely within the pelvic fracture group, where seven out of sixteen Type C patients died, generating an in-hospital mortality rate of 43.8 percent, alongside a single death among the twenty-eight Type A patients, representing 3.6 percent, and zero deaths within Types B and D. Renewed arterial hemorrhage originating from the previously treated inferior epigastric artery territory was documented in three patients, representing an overall rebleeding rate of 4.8 percent; notably, all three rebleeding events occurred exclusively among patients within the Type C category, yielding a category-specific rebleeding rate of 18.8 percent.

To investigate independent factors underlying the composite clinical failure endpoint, the investigators conducted exploratory logistic regression analyses among fifty-nine patients, excluding the three Type D cases due to complete separation arising from zero adverse events. In univariable evaluations, adverse outcomes demonstrated suggestive associations with the Type C category, disseminated intravascular coagulation, systolic blood pressure below 80 millimeters of mercury, serum lactate of 4 millimoles per liter or higher, transfusion of ten or more red blood cell units within twenty-four hours, and polytrauma. Polytrauma was documented in ten patients, all confined entirely to Type C. To preserve model stability and avoid collinearity with shock indices, the multivariable regression model retained three key nonredundant predictors: the Type C category, disseminated intravascular coagulation, and profound initial hypotension.

The final multivariable logistic regression model revealed that the Type C category was independently associated with the composite clinical failure endpoint, yielding an adjusted odds ratio of 4.6 with a ninety-five percent confidence interval ranging from 1.3 to 16.2. The presence of disseminated intravascular coagulation, diagnosed according to Japanese Ministry of Health, Labour and Welfare criteria and documented in 41.9 percent of the overall cohort, was also independently associated with failure, showing an adjusted odds ratio of 3.7 with a ninety-five percent confidence interval of 1.1 to 12.3. Severe hypotension, defined by a pre-procedural systolic blood pressure below 80 millimeters of mercury and recorded in 35.5 percent of patients, demonstrated an adjusted odds ratio of 3.1 with a ninety-five percent confidence interval of 1.0 to 9.7. From these factors, the researchers constructed an exploratory descriptive scoring system assigning two points for Type C, two points for disseminated intravascular coagulation, and one point for systolic blood pressure below 80 millimeters of mercury, which correlated with an escalating clinical failure rate rising from 7.1 percent in low-risk scores to 71.4 percent among scores of four to five.

The authors emphasized several critical limitations, noting that their retrospective, single-center design, limited sample size, and non-randomized embolic material selection require cautious interpretation. Furthermore, sensitivity analyses substituting clinical etiology for clinico-angiographic categories yielded qualitatively similar findings, confirming that the framework functions not as an etiology-independent causal determinant of mortality, but as an operational roadmap to identify retrograde collateral flow and guide comprehensive embolization planning in acute hemorrhage.

Subject of Research: Transcatheter arterial embolization outcomes and collateral vascular pathways in inferior epigastric artery hemorrhage

Article Title: Collateral-pathway–oriented clinico-angiographic framework for inferior epigastric artery–related hemorrhage: embolization planning and clinical outcomes

Article References: Tokue, H., Ebara, M., Tokue, A., & Tsushima, Y. (2026). Collateral-pathway–oriented clinico-angiographic framework for inferior epigastric artery–related hemorrhage: embolization planning and clinical outcomes. CVIR Endovascular, 9(1), Article 91. https://doi.org/10.1186/s42155-026-00742-3

Image Credits: AI Generated

DOI: 10.1186/s42155-026-00742-3

Keywords: transcatheter arterial embolization, inferior epigastric artery, interventional radiology, pelvic fracture, postpartum hemorrhage, corona mortis, rectus sheath hematoma, disseminated intravascular coagulation, Collateral-pathway, oriented, clinico-angiographic, framework