Abstract
Background
Gastroepiploic artery aneurysms (GEAAs) and their rupture are very rare but often serious. However, emergency management has yet to be standardized.
Purpose
To clarify the clinical features of GEAAs and outcomes of transcatheter arterial embolization (TAE).
Materials and Methods
This is a retrospective, single-center 12 case series of GEAAs experienced between 2006 and 2023. We reviewed medical records to determine the case background, angiographic images, TAE techniques and success rate, and outcomes. Abdominal angiography was performed via the femoral artery to identify the inflow and outflow vessels of the GEAAs. Subsequently, a microcatheter was advanced to the target site, and embolization was performed with microcoils and/or NBCA-Lipiodol mixture.
Results
Nine ruptured and one unruptured pseudo-GEAA cases were treated by emergency TAE, while the remaining two unruptured cases were treated electively. The average diameter of the ruptured GEAAs was 7.9 mm. The most common underlying diseases were segmental arterial mediolysis in 4 cases. Morphological classification revealed 5 cases of dissecting, 4 of pseudo, and 3 of true. TAE was successful in seven of the nine ruptured and in all three unruptured cases. Two patients with ruptured GEAAs after unsuccessful TAE were subsequently saved by surgery. TAE using the triple coaxial catheter system was performed in 7 cases with good results.
Conclusions
Even small-diameter GEAAs can rupture, resulting in life-threatening conditions, but emergency TAE is safe and effective. However, there are some cases in which TAE fails, so it is important to make a prompt decision to proceed to surgical treatment.
Keywords
Introduction
Abdominal visceral artery aneurysms (VAAs) are rare with a prevalence of 0.1%−2% 1 ; gastroepiploic artery aneurysms (GEAAs) are even rarer, accounting for approximately 0.4% of VAAs. 2 However, GEAAs are at high risk of rupture, and the American Society for Vascular Surgery guidelines 3 recommend medical intervention for GEAAs regardless of their size. GEAAs are rare, so only single case reports have been published and few comprehensive case studies have been published. 4 We experienced 12 cases of GEAAs and treated them with transcatheter arterial embolization (TAE). Here, we report a retrospective case series on the clinical features and treatment outcomes of these 12 cases.
Patients & methods
This study included 12 cases of GEAAs experienced at our hospital between 2006 and 2023. We retrospectively reviewed medical records to determine the case background, angiographic images, the method and success of TAE, postoperative complications and outcomes.
Technical success was defined as the complete disappearance of blood flow to the aneurysm on the digital subtraction angiography after TAE. Clinical success was defined as no 30-day rebleeding and mortality after TAE and no additional surgery or re-TAE.
The diagnosis of GEAAs was made using contrast enhanced computed tomography (CT). Cases of ruptured GEAAs and pseudoaneurysms were treated as soon as possible as an emergency by TAE, while cases of unruptured GEAAs were treated electively. The abdominal angiography procedures were performed by two hepatologists having ample experience in endovascular treatment with the assistance of young clinicians including senior residents. All endovascular treatments were performed with sufficient explanation to each patient and their written consent.
The femoral artery was punctured using the Seldinger’s technique, and a 5 Fr-25 cm long sheath introducer (Radifocus® Introducer II-H, Terumo, Tokyo, Japan) was inserted into the abdominal aorta. A 5 Fr Shepherd hook-type angiographic catheter (Goodtec® angiographic catheter, Goodman, Nagoya, Japan) was used to visualize the celiac artery and other areas to confirm the location of the aneurysm. After confirming the location of the GEAAs, we selectively advanced a 1.98 Fr-125 cm long microcatheter (Parkway SOFT®; Asahi Intecc, Seto, Japan) using the guidewire (Radifocus GT® wire 0.016 inch-150 cm long, Terumo, Tokyo, Japan, or Labyrinth® 0.014 inch-180 cm long, PIOLAX, Yokohama, Japan) and embolized the GEAAs using the isolation and/or packing technique. Interlock® 0.016-inch detachable microcoils (Boston Scientific, Marlborough, MA, USA) and Tornado® Embolization Coil (Cook Medical, Bloomington, IN, USA) were used for coil embolization. TAE was also performed using a mixture of n-butyl-2-cyanoacrylate (NBCA: Histoacryl®, B. Braun, Melsungen, Germany) and Lipiodol® 480 (Guerbet, Villepinte, France) in a 1:1 to 1:3 ratio (50%–20%) as glue. There are no clear criteria for selecting the embolic material, but if coil embolization is inappropriate, that is, if the parent artery of the pseudoaneurysm cannot be completely embolized with microcoils, or if the catheter cannot reach the target aneurysm, then NBCA-Lipiodol glue is injected into the aneurysm to fill it. When the aneurysm was too far away to be reached by the coaxial approach, a high-flow microcatheter (Masters HF® 2.6 Fr-125 cm long: Asahi Intecc) was used as the second catheter, and a thin non-tapered type 1.9 Fr-150 cm long microcatheter (Carnelian® Marvel Non-Taper, Tokai Medical Products, Kasugai, Japan) as the third catheter was then passed through it to reach the target aneurysm (triple-coaxial catheter system: TCCS).5–7 TCCS was used when it was determined that the target aneurysm could not be reached with a simple coaxial system (only one microcatheter) because it was not only far from the parent catheter but also had many bends in the artery. A second high-flow microcatheter provides backup by overcoming several bends, and further facilitates the advancement of a third coaxial catheter to the target aneurysm. In the TCCS method, NBCA-Lipiodol glue was used as the embolic material.
Perioperative management
In cases of ruptured GEAA, sufficient intravenous administration of Ringer’s solution and optimal blood transfusions were performed to stabilize hemodynamics, followed by prompt transition to endovascular treatment. Intravenous antibiotics (cefazolin) were administered immediately before endovascular treatment and for approximately 2 days afterward. The patients were discharged after observation periods to check for complications and after their overall conditions, including other illnesses, had improved. Generally, we do not provide long-term follow-up after discharge.
Ethical approval
This retrospective observational case series was conducted in accordance with the ethical guidelines of the Declaration of Helsinki and was approved by the Institutional Review Board of our institution with approval number: 23-047 in 2023. Given the retrospective design and use of anonymized clinical data, the requirement for obtaining written informed consent from each patient was waived. Instead, patients were provided the opportunity to decline participation through an opt-out process, whereby study information was made publicly available on the institutional website.
Results
Background and clinical features of GEAA cases
Profiles of patients with gastroepiploic artery aneurysms.
CVD: cerebrovascular disease, IHD: ischemic heart disease, HT: hypertension, HL: hyperlipidemia, DM: diabetes mellitus, HD: hemodialysis for chronic renal failure, M: male, F: female.
Status at first visit in patients with ruptured gastroepiploic artery aneurysm.
SBP: systolic blood pressure, HR: heart rate, Hb: hemoglobin, RBC: red blood cells, SD: standard deviation.
Abdominal angiography
Findings of gastroepiploic artery aneurysms and outcomes of TAE.
TAE: transcatheter arterial embolization, App.: approach, TCSS: triple coaxial catheter system, DA: dissecting aneurysm, PA: pseudoaneurysm, TA: true aneurysm, SAM: segmental arterial mediolysis, DU: duodenal ulcer, ANCA vasculitis: antineutrophil cytoplasmic antibody associated vasculitis, R: via the right gastroepiploic artery, L: via the left gastroepiploic artery, NBCA: n-butyl-2-cyanoacrylate and lipiodol mixture, MC: microcoils.
aCase 3 was saved by the surgeons performing distal pancreatectomy, splenectomy, and partial omentectomy.
bCase 8 was saved by the surgeons performing partial omentectomy.
Transcatheter arterial embolization (TAE)
TAE was technically successful in seven of the nine ruptured cases (77.8%) and in all three unruptured cases. The procedure time was 64.4 ± 9.2 min in ruptured GEAA cases and 75.0 ± 33.6 min in all cases. All patients who had successful TAE recovered and were discharged. Excluding one patient who required long-term hospitalization for steroid treatment for ANCA-associated vasculitis, the average length of hospital stay after TAE was 12.3 days. Two patients with ruptured GEAAs after unsuccessful TAE were subsequently saved by surgery. These 2 cases of unsuccessful TAE are shown in Fig. 1. A 76-year-old woman was transferred to the hospital with chest pain and hemorrhagic shock. CT scan revealed a diagnosis of ruptured left GEAAs, and emergency angiography was performed. In addition to the left GEAAs, multiple bead-like dilated aneurysms were also found in the right gastroepiploic artery, and the patient was clinically diagnosed with SAM. TAE of the multiple left GEAAs was abandoned, and the patient was saved by the surgeons performing distal pancreatectomy, splenectomy and partial omentectomy. In the resected specimen, pathological findings of SAM were observed in the ruptured GEAA. The other case was a 65-year-old man undergoing hemodialysis for chronic renal failure who was transported to our hospital with sudden epigastralgia and hemorrhagic shock. Contrast-enhanced CT showed an omental hematoma and a ruptured right GEAA, so emergency angiography was performed. Right gastroepiploic arteriography revealed a pseudoaneurysm, but the contrast injection pressure alone caused rupture of another arterial branches not touched by the device, resulting in extravasation of the contrast medium. Further selective cannulation of the microcatheter became difficult, so NBCA-Lipiodol glue was injected, however, it only reached part of the GEAA, resulting in only proximal embolization. Because the TAE was deemed incomplete, the surgeons performed a partial omentectomy and saved the patient. Pathological examination of the resected specimen revealed a diagnosis of SAM. The technical and clinical success rates of TAE were both 83.3% (10 of 12 cases), and there were no rebleeding and deaths. Two cases of unsuccessful transcatheter arterial embolization (Case 3 & 8, Tables 1, 2, 3). A 76-year-old woman (Case 3) was transferred with sudden chest pain and hemorrhagic shock. (a) Celiac arteriography showed multiple bead-like dilated aneurysms not only in the left gastroepiploic artery (white arrow) but also in right gastroepiploic artery (black arrow). (b) Splenic arteriography demonstrated multiple bead-like dilated aneurysms in the left gastroepiploic artery (white arrow) and a saccularly expanded aneurysm (black arrowhead) in the periphery. We determined that it would be difficult to reach the target site with the catheter and abandoned endovascular treatment. A 65-year-old man (Case 8) undergoing dialysis for chronic renal failure was brought to the hospital with sudden upper abdominal pain and hemorrhagic shock. (c) Celiac arteriography showed an aneurysm at the periphery of the right gastroepiploic artery (black arrowhead). (d) When right gastroepiploic arteriography was performed (tip of the microcatheter: white arrow), the aneurysm (black arrowhead) was visualized, but the pressure of the contrast medium alone caused rupture of an artery not touched by the device, resulting in extravascular leakage (into the omentum) of the contrast medium (white arrowhead). Because we were unable to reach the target aneurysm with the catheter, we injected NBCA-Lipiodol from a distance, but this resulted in only proximal embolization.
Complications
Of the seven patients who underwent upper gastrointestinal endoscopy after TAE, two (Case 6 and 12) were found to have gastric ulcers. These ulcers were free of bleeding or exposed vessels at the ulcer base, and improved with oral medication. No other significant treatment-related complications were observed.
Presentation of typical images
Typical angiographic images of dissecting aneurysm are shown in Fig. 2. A 67-year-old man presented with abdominal pain and was diagnosed with omental hematoma due to ruptured GEAAs on abdominal CT. There were multiple irregular lead-pipe-like vascular dilations in the left gastroepiploic artery, which were successfully treated by TAE using a NBCA-Lipiodol glue by the TCCS method. Case of ruptured dissected left gastroepiploic artery aneurysm (Case 7, Tables 1, 2, 3) A 67-year-old man was transferred to our hospital with abdominal pain. (a) Contrast-enhanced computed tomography revealed a hematoma on the left side of the stomach, with dilated vascular structures inside (white arrowhead). (b) Celiac arteriography revealed irregular, lead-pipe-like dilation of the left gastroepiploic artery (black arrowhead). (c) Left gastroepiploic arteriography using a triple coaxial catheter system (black arrow indicates the tip of the second catheter, and white arrow indicates the tip of the third catheter) revealed multiple dissecting aneurysms (black arrowheads), and transcatheter arterial embolization was then performed using NBCA-Lipiodol mixture. (d) After treatment, celiac arteriography showed that the left gastroepiploic artery was embolized like the casting mold (black arrowhead).
Typical angiographic images of pseudoaneurysm are shown in Fig. 3. A 58-year-old man was found to have an aneurysm on the ventral side of the greater curvature of the stomach during follow-up CT scan for ANCA-associated vasculitis. Right gastroepiploic arteriography revealed an irregular sac-like appearance slightly delayed from the early arterial phase, and a pseudoaneurysm of the gastroepiploic artery was diagnosed. The catheter was delivered into the GEAA using the TCCS method, and TAE was successfully performed by injecting NBCA-Lipiodol glue. Case of gastroepiploic artery pseudoaneurysm (Case 12, Tables 1 and 3). A 58-year-old man underwent a contrast-enhanced computed tomography for follow-up of antineutrophil cytoplasmic antibody-associated vasculitis. (a) Axial section, (b) coronal section. A pseudoaneurysm was noted on the ventral side of the greater curvature of the stomach (white arrow). (c) Right gastroepiploic arteriography revealed a pseudoaneurysm that was visualized slightly later than the early arterial phase. (d) The triple coaxial catheter system (black arrow indicates the tip of the second catheter, and white arrow indicates the tip of the third catheter) was used to reach the gastroepiploic artery aneurysm. (e) Transcatheter arterial embolization was performed using NBCA-Lipiodol mixture. (f) After treatment, right gastroepiploic arteriography showed that the aneurysm was embolized like the casting mold (black arrowhead).
Typical angiographic images of true aneurysm are shown in Fig. 4. A 79-year-old woman was found to have an aneurysm near the pancreas head on a follow-up abdominal CT scan for another disease. Celiac arteriography revealed a spherical aneurysm at the root of the right gastroepiploic artery from the early arterial phase, which was diagnosed as a true aneurysm. The outflow tract of the GEAA was embolized with microcoils (distal isolation), and the inflow tract from the GEAA was embolized with NBCA-Lipiodol glue (packing and proximal isolation). Case of true aneurysm of the right gastroepiploic artery (Case 10, Tables 1 and 3). A 79-year-old woman was referred to our hospital because a screening computed tomography at another hospital revealed an aneurysm near the pancreas head. (a) Axial section, (b) 3-dimensional reconstruction imaging of computed tomography (RAO 30°). A spherical aneurysm (white arrow) was noted on the ventral side of the pancreas head. (c) Celiac arteriography demonstrated a spherical aneurysm in the proximal right gastroepiploic artery from the early arterial phase (black arrowhead). (d) First, selective embolization using microcoils was performed on the right gastroepiploic artery and the anterior superior pancreaticoduodenal artery, which serve as the outflow tracts of the aneurysm. (e) The aneurysm was embolized with NBCA-Lipiodol mixture. (f) After treatment, celiac arteriography showed that the aneurysm was embolized like the casting mold (black arrowhead).
Discussion
Abdominal VAAs are a less common condition. 1 The most common location of aneurysms is the splenic artery (60%), followed by the hepatic artery (20%), superior mesenteric artery (5.5%), and celiac artery (4%). 2 Gastric artery aneurysms and GEAAs account for only 4% of all abdominal VAAs, with GEAAs being approximately one-tenth as rare. 2 It has been reported that right GEAAs are more common than left GEAAs,9,10 but where is the boundary between the left and right gastroepiploic arteries? In general, it is difficult to define the center of the arteries that form the arcade. The position of the center of the gastroepiploic artery varies depending on the stomach and other intestinal contents, therefore there are some GEAAs in which it is difficult to define the left or right. We believe it is reasonable to classify the aneurysm as left or right based on whether the catheter was advanced via the right or left gastroepiploic artery on angiography. The frequency of approach (Table 3) was via the right in 7 cases and via the left gastroepiploic artery in 5. However, when limited to the ruptured GEAAs cases only, the situation was reversed, with 4 cases via the right and 5 cases via the left gastroepiploic artery. Since the left GEA has a small diameter, it may be easy to tear with slight expansion. The causes of GEAAs 11 include atherosclerosis, trauma, iatrogenic,12,13 vasculitis, 14 collagen diseases, congenital anomalies, infection, and SAM.7,15 Among our 12 GEAA cases, SAM was the most common, with four cases, two of which were pathologically confirmed. Three of these SAM cases showed typical multiple aneurysms with beaded dilation, and dissecting aneurysm was the predominant type. Of the four cases of pseudoaneurysm, the cause was clear in two cases: duodenal ulcer in one case and ANCA-associated vasculitis in the other case. Although there is no evidence, the possibility that arterial vascular fragility is related to aneurysm formation should be considered, as three of the 12 GEAA cases were on hemodialysis.
Classification of aneurysms into true aneurysm, pseudoaneurysm, and dissecting aneurysm is based on pathological diagnosis, and it may be difficult to determine the type based on imaging alone. Although there are no comprehensive reference books, we know from experience the typical angiographic findings of each aneurysm type. True aneurysms have a spherical to spindle-like shape, and reflecting the enlargement of the vascular lumen of a single artery, angiography depicts the inflow artery, aneurysm, and outflow artery without delay. Pseudoaneurysms are cavities filled with blood outside the arterial structure, so they are often irregular shape and there is often delayed visualization of the aneurysm and outflow from the aneurysm. Dissecting aneurysms can be diagnosed with certainty if the intimal flap is visible; however, it is often not visible. A lead pipe-like dilation or delayed outflow of contrast medium from the false lumen is often seen, which are supporting findings for the diagnosis of dissecting aneurysm. When classified based on these angiographic findings, there were 3 cases of true aneurysms, four of pseudo, and five of dissecting, with dissecting aneurysms being the slightly more common.
It has been reported that rupture bleeding accounts for 90% of GEAAs diagnosis, and 70% of cases are fatal. 1 Of our 12 GEAA cases, nine cases were ruptured. Although a number of our cases is small, the aneurysm diameter was smaller in ruptured cases compared with elective treatment cases, which supports the indication of treatment regardless of aneurysm diameter, as stated in the guideline. 3 While there is no established theory in the literature, GEAAs may be prone to rupture even if they are small in diameter, given that they occur due to the dilation of blood vessels that are originally small in diameter, and due to vascular fragility caused by various diseases.
Rupture of GEAAs almost always results in intraperitoneal hemorrhage and/or intraomental hematoma, which is demonstrated by CT. The arterial phase of dynamic CT with rapid bolus injection of contrast is optimal for diagnosing GEAA. However, when patients visit the emergency room for abdominal pain or other conditions, a standard single-phase contrast CT scan is often performed to broadly screen for various differential diagnoses. Therefore, the imaging methods used in CT scans to diagnose GEAA are not standardized in this case series, which is unique to the real world.
There are cases in which GEAAs perforate the gastrointestinal tract, and although this is a rare condition, it is an important differential disease as a cause of gastrointestinal bleeding. We also experienced a case of hematemesis due to GEAA rupture caused by a duodenal ulcer. The patient was resistant to endoscopic hemostasis, but was saved by TAE.
Treatment of GEAAs includes open surgery,16–18 laparoscopic surgery,10,19 and endovascular treatment.4,7,20,21 In patients with unstable vital signs, laparotomy is the treatment of the first choice.16,18 Although it is difficult to remove the hematoma and identify the bleeding vessels, ligation of the responsible vessels or partial omental resection provides curative treatment. In recent years, TAE based on angiography, which is superior in identifying aneurysms and bleeding sites, has become the mainstream4,7,15 and is recommended in the guidelines (Level 1-B). 3
The right gastroepiploic artery has a relatively large diameter and is easy to approach with a microcatheter, but if GEAAs form at its periphery, the microcatheter must be advanced a considerable distance. Furthermore, the left gastroepiploic artery typically branches off from the splenic artery, is small in diameter, and is located quite far from the celiac artery, making it difficult to advance a microcatheter. A coaxial system consisting of a four to five Fr catheter plus approximately 2 Fr microcatheter may not be able to reach the GEAAs sufficiently, making it difficult to isolate the GEAAs using microcoils. We have also observed a case where blood vessels not reached by guidewires or microcatheters ruptured with contrast injection alone (Case 8, Fig. 1 (c) and (d)), and similar reports have been published. 22 Therefore, in cases involving vascular fragility, such as SAM, vasculitis, and hemodialysis, it is important to be extremely gentle when operating on arteries that do not have aneurysms. Although it is possible to embolize the aneurysm sufficiently by injecting glue such as NBCA-Lipiodol glue even at a location quite proximal to the aneurysm, the glue may become stuck before reaching the aneurysm, resulting in proximal embolization (our Case 8, Fig. 1 (c) and (d)). Therefore, TCCS,5,7 which allows the microcatheter to reach more distally in a stable manner, may be useful. We have recently used the TCCS method in seven cases, successfully reaching the GEAAs and performing TAE using glue. For far right GEAAs and left GEAAs, we will use this TCSS method as our first choice.
Nozawa Y et al. performed TAE on 15 cases of GEAAs and reported that no obvious development of omental infarction or necrosis was observed. In our 12 cases, no omental infarction or necrosis was observed, but gastric ulcers were observed in two cases. TAE of GEAAs achieved a high success rate of 3/3 (100%) in elective cases and 7/9 (77.8%) in emergent cases, and was evaluated as a useful treatment method without any serious complications. In addition, the omentum is anatomically the most ventral part, and surgery can be performed without damaging important organs. Therefore, in cases that blood circulation cannot be maintained, or that the catheter cannot be advanced to the treatment site of the target aneurysm, the treatment should be switched from endovascular treatment to surgery as early as possible.
This study spans a long period of 18 years, but only 12 cases of GEAA were observed, indicating its rarity. While there may be some temporal bias due to advancements in the operators’ techniques over time, there were no changes in the primary operators, and therefore the impact on the outcomes is considered minimal. Although imaging techniques such as CT scans have become higher resolution, aneurysm detection was comparable to that in 2006, and catheters and embolizing materials have remained largely unchanged. Although the circumstances of each of the 12 cases may have differed slightly over the 18 years, we believe they can be treated similarly within a single study.
This study has several limitations. First, it is a retrospective, single-center experience. Second, it included only a small number of patients despite the very long study period because GEAA is a rare disease. Third, selection bias exists regarding patients and treatment methods. Fourth, because this is a retrospective study of real-world data, CT imaging methods and post-treatment follow-up were not uniform, and there was no long-term follow-up.
In conclusion, even small-diameter GEAAs can rupture, resulting in life-threatening conditions, but emergency TAE is safe and highly effective. However, there are some cases in which TAE fails due to vascular fragility or anatomical difficulties, so it is important to make a prompt decision to proceed to surgical treatment.
Footnotes
Acknowledgments
Ethical considerations
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This retrospective observational study received ethical approval by the Institutional Review Board (IRB) of our hospital (23-047).
Consent for publication
Since this study has removed any personally identifiable information, consent for publication is generally not required. The study received IRB approval, and the need for informed consent was waived with only opt-out disclosure. Of course, written informed consent for examinations and treatments was obtained in each individual case.
Author contributions
Nobuo Waguri and Akihiko Osaki contributed to the study conception and design. Angiography and TAE were performed by Nobuo Waguri, Akihiko Osaki, and Takashi Owaki. The clinical data was collected by Nobuo Waguri, Junji Kohisa, Kennichi Takaku and Munehiro Sato. The manuscript was written by Nobuo Waguri and was reviewed by Takashi Owaki and Akihiko Osaki. All authors read and approved the final manuscript.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
Not applicable to protect patient privacy.
