Endoscopic ultrasound-guided coil embolization for gastric varices: A promising alternative to traditional therapies

Abstract

Endoscopic ultrasound (EUS) guided vascular interventions have expanded the reach of therapeutic endoscopy to include vascular pathology previously inaccessible by endoscopists. Gastric variceal bleeding comprises 20% of all variceal bleeding and is associated with high morbidity and mortality. Historically, endoscopic injection of thrombosis-inducing agents such as glue has been used. However, glue injection carries potential risks including systemic embolization, damage to the endoscope, and recurrent bleeding. The introduction of hemostatic coils has revolutionized the endoscopic approach, with EUS-guided coil embolization emerging as an effective and safe modality for the management of gastric varices (GVs). When compared with conventional glue injection, EUS-guided embolization is associated with improved visualization, higher efficacy, and better safety profile. Despite its expanding adoption, the standardization of EUS guided embolization remains a challenge. High-quality studies are needed to standardize this promising technique and define its role in clinical practice. In this review, we will discuss the indications, efficacy, techniques, and various approaches for EUS-guided embolization of GVs.

Key Words: Gastric variceal hemorrhage; Endoscopic ultrasound; Therapeutic endoscopy; Coil embolization; Cyanoacrylate; Gastrointestinal bleeding

Core Tip: Endoscopic ultrasound (EUS)-guided vascular interventions have changed the treatment landscape of gastric varices, providing a safer, more effective alternative to conventional glue injection. EUS-guided coil embolization improves visualization, provides higher efficacy and is associated with fewer complications including systemic embolization and recurrent bleeding. Nevertheless, the procedure remains difficult to standardize, and high-quality data is lacking to inform and optimize the treatment approach. This review will discuss the indications, techniques and emerging role of EUS-guided embolization in the management of gastric variceal hemorrhage.

INTRODUCTION

Gastric varices (GV) are a common cause of upper intestinal bleeding. They occur in up to 20% of patients with portal hypertension; a common sequelae of liver cirrhosis[1]. Although less frequent compared to esophageal variceal (EV) hemorrhage, gastric variceal hemorrhage (GVH) carries a higher risk of rebleeding and mortality[2]. The most widely accepted classification of GVs is the Sarin classification which divides GVs into four types based on their relationship to EV and/or their anatomical location[3]. Gastroesophageal varices (GOV) make up to 96% of all GVs which are divided into GOV1 (continuation of EV along the lesser curvature of the stomach) and GOV2 (continuation of EV extending to the fundus of the stomach). Isolated gastric varices (IGV) carry a higher risk of bleeding compared to GOV and are divided into IGV1 (varices located in the fundus only without esophageal involvement) and IGV2 (varices located in other parts of the stomach without esophageal involvement)[2,4]. In fact, the size rather than the type of the varix remains the most important indicator for bleeding[2]. The Sarin classification helps determine the appropriate treatment strategy for GV. GOV1 are managed similarly to EV typically with banding, while cardio-fundal varices (GOV2 and IGV1) require more specialized interventions.

In 1986, Soehendra et al[5] were the first to describe the treatment of GVs through endoscopic injection of thrombosis-inducing agents such as cyanoacrylate glue (CYA). This approach became a mainstay in GV management. Although it is relatively safe, this technique carries the risk of distal embolization caused by downstream flow of glue before polymerization[6].

More recently, endoscopic ultrasound (EUS)-guided vascular intervention emerged as a promising approach in the management of GVs. Under EUS guidance, different hemostatic adhesives and devices can be injected into GV including CYA alone, coils alone, coils with CYA, thrombin, and coils with absorbable gelatin sponge (AGS)[6]. Nonetheless, standardizing EUS guided embolization remains challenging due to the scarcity of high-quality data guiding the optimal management strategy. In this review, we will discuss the indications, efficacy, techniques, complications, and various approaches for EUS-guided coil embolization (EUS-coil) of GVs.

LITERATURE SEARCH

The literature search was conducted using MEDLINE and EMBASE electronic databases up to May 2025. A combination of text words and MeSH terms (gastrointestinal hemorrhage, liver cirrhosis) for the relevant topics were used in combination with GVs, embolization, and coils. Cross-references were identified manually through the citation list of selected articles to capture additional resources. Our search yielded 182 results published between 1979 and 2025. Relevant articles were reviewed, and additional papers were captured through the citation list of reviewed articles. As this is a narrative review, no ethical approval was required.

TRADITIONAL APPROACH TO GVH

The general principles for the management of GVH are similar to EV bleeding including volume resuscitation, correction of coagulopathy, and pharmacological interventions such as prophylactic antibiotics and vasoactive drugs[7]. In addition, the endoscopic management of GOV1 and EV is comparable. However, the management of cardio-fundal varices which includes IGV1 and GOV2 is less clear. For primary prophylaxis of cardio-fundal variceal bleeding, the American Association for the Study of Liver Diseases (AASLD) recommends non-selective beta-blockers (NSBB)[8]. For bleeding cardio-fundal varices, Trans-jugular Intrahepatic Portosystemic Shunt (TIPS) and Balloon-Occluded Retrograde Transvenous Obliteration (BRTO) have been traditionally considered first-line. BRTO requires a gastro-renal shunt to be present but might increase portal pressures while TIPS diverts the portal flow and increases the risk of hepatic encephalopathy[9,10]. Both TIPS and BRTO are highly effective for hemostasis of bleeding GVs with more than 90% efficacy[11]. It is noteworthy that some studies consider TIPS to be less effective for GVs than for EV[10]. Since BRTO increases portal hypertension, a hepatic venous portal gradient of 12 mmHg is an indication to use TIPS rather than BRTO[8,10,12].

Despite the high efficacy of TIPS and BRTO, TIPS can be complicated by hepatic encephalopathy and TIPS dysfunction resulting in rebleeding, while the increase in portal pressures following BRTO may worsen underlying ascites[13,14]. As an alternative, direct endoscopic injection of tissue adhesives has been proposed. Two main classes of adhesives have been evaluated including synthetic glues and biologic adhesives such as thrombin[15]. The majority of available data is based on endoscopic injection of synthetic glue into the GVs, mainly N-butyl-2-cyanoacrylate (NBCA), and 2-octyl-cyanoacrylate. NBCA has a faster polymerization time and is often mixed with lipiodol to prevent premature solidification and adherence to equipment[16].

Mishra et al[17] showed a significantly lower risk of rebleeding for secondary prophylaxis of cardio-fundal varices with CYA injection compared to beta-blockers (15% vs 55%, P = 0.004). In their meta-analysis, Qiao et al[18] found that 93.9% of actively bleeding GVs were controlled with CYA glue compared to 79.5% with endoscopic band ligation. However, in centers where CYA is unavailable, endoscopic band ligation is a good alternative nonetheless[9]. However, when compared with TIPS in cirrhotic patients with acute bleeding from GVs, CYA glue injection was less effective in preventing rebleeding (38% vs 11%, P = 0.014) with similar survival and frequency of complications[19].

While direct endoscopic injection of synthetic glue is relatively safe, several concerns have been raised related to distal embolization of the glue before polymerization resulting in pulmonary embolism, splenic infarcts, septic emboli, stroke, and coronary emboli[2]. Transient abdominal pain and fevers have also been reported. In addition, a frequent concern with CYA glue injection is instrument damage caused by polymerization within the working channel and adherence to the scope tip[20]. Given the limited data and potential complications, endoscopic therapy with direct glue injection is not advised for management of GV. The AASLD guidelines suggest that CYA injection is an option for cases in which TIPS is not feasible. Additionally, according to the Baveno VI consensus, more studies are required to better assess the risk–benefit profile of CYA[8,9].

EUS FOR THE MANAGEMENT OF GVS

The introduction of EUS for the management of GV has garnered significant interest over the past few years with several advantages over direct endoscopic injection for both diagnostic and therapeutic indications. First, EUS optimizes visualization of GV compared to standard upper endoscopy especially in the setting of active bleeding, with better visualization of the gastric wall and associated vasculature[21]. In addition, confirmation of variceal obliteration remains subjective with direct glue injection and relies on determining ‘hardening’ of the varix post embolization which may lead to excessive glue injection in case of uncertainly increasing the risk of distal embolization[22]. Alternatively, EUS can effectively confirm variceal obliteration by real-time visualization of Doppler showing the absence of flow within the varix. In fact, EUS allows the injection of different adhesives and devices into the GV including CYA, coils, thrombin, AGS while direct endoscopic injection was mainly limited to glue injection.

The use of EUS-guided cyanoacrylate injection (EUS-CYA) was first described by Romero-Castro et al[23] in 2007 in a series of 5 patients, followed up for 10 months, for primary prophylaxis of GV bleeding. Interestingly, none of them experienced any complications or re-bleeding during that period. Since, several studies have described the efficacy of EUS-CYA for both primary and secondary prophylaxis for GV bleeding[24-27] (Table 1). However, EUS-CYA has a comparable adverse effect profile to direct endoscopic injection despite the injection of less CYA under EUS guidance including abdominal pain, fever, and transient bacteremia[28]. Romero-Castro et al[24] reported a 47% rate of systemic embolization of glue with pulmonary embolism noted on routine post-procedure imaging, although all patients were asymptomatic. Other studies have reported a 2%-6% risk of pulmonary embolism and splenic infarcts[27,28].

Table 1 Studies evaluating endoscopic ultrasound-guided coil embolization, endoscopic ultrasound-guided cyanoacrylate injection, or a combination of both evaluating their efficacy in variceal obliteration, risk of rebleeding, and adverse events since 2022.

Intervention	Year	Ref.	Study type	Number of patients (n)	Variceal obliteration (%)	Re-bleeding (%)	Adverse events1 (%)
Coils	2020	Robles-Medranda et al[38]	Single-center RCT	30	90	26.9	0
Coils + CYA	2020	Robles-Medranda et al[38]	Single-center RCT	30	100	3.4	0
Coils + CYA	2021	Kouanda et al[36]	Retrospective analysis	62	96.7	3.2	2.5
Coils + CYA	2022	Baig et al[47]	Meta-analysis	323	96.76	5.09	2.2
Coils + CYA	2024	Chen et al[48]	Retrospective analysis	21	90.5	4.8	-
CYA	2024	Chen et al[48]	Retrospective analysis	36	70.6	27.8	-
Coils2	2024	Bazarbashi et al[46]	Retrospective analysis	106	73.6	15.2	1.8

¹Only major adverse events such as systemic embolization were accounted for.

²82% of patients required adjunctive glue or absorbable gelatin sponge injection.

RCT: Randomized control trial; CYA: Cyanoacrylate.

Given the adverse event profile of EUS-CYA, EUS guided coil injection with or without glue was introduced to minimize the risk of distal embolization. The coil, by retaining glue within the varix, minimizes that risk[29]. This was first described by Romero-Castro et al[30] in 2010 with effective eradication of GV in 3 out of 4 patients.

APPROACH AND TECHNIQUE

There is no standardized approach for the EUS-coil and multiple variations have been described. Below we will describe our approach for EUS-coil. EUS-guided vascular therapy is a multistep approach requiring a multitude of instruments and accessories, including a linear array echoendoscope although some have used a forward-viewing linear array echoendoscope[29,31], a 19 or 22-gauge fine-needle aspiration (FNA) needle, several syringes filled with normal saline or water, coils (various sizes, lengths, and types of coils exist), and glue (CYA). Fluoroscopic guidance may be helpful to visualize the deployment of coils but is not necessary.

PRE-EMBOLIZATION PHASE

We recommend this procedure to be performed under general anesthesia with endotracheal intubation given potential but small risk of bleeding following the puncture of varices. We give one dose of antibiotics prior to the procedure for prophylaxis. Two units of packed red blood cells are made available in case significant bleeding is encountered. The procedure is usually performed with the patient in left lateral position; however, position change may be helpful to optimize visualization in case significant bleeding is encountered.

The echoendoscope is then positioned either in the distal esophagus for a trans-crural approach or in the stomach for a trans-gastric approach. In case of difficult visualization of the intended target, the gastric fundus may be filled with water to enhance the endosonographic image. It is crucial at this point to examine the submucosal GVs, the intended target for embolization, and differentiate them from the peri-GVs. Adequate examination also allows the endosonographer to detect the feeder vessel during its passage into the submucosal space. Measure the maximal axial diameter as coil size and length is selected based on the diameter of the target varix. Coil diameter can range anywhere from 5 mm to 20 mm. Typically, coils need to be 20% to 30% larger than the diameter of the varix. A 19-gauge or a 22-gauge FNA needle can be used to target the feeder vessel or the varix itself and the coils are introduced using the stylet as a pusher. We prefer shorter length coils as longer ones occasionally coil within the FNA needle during advancement. 19-gauge needles are compatible with 0.035-inch coils and 22-gauge needles are compatible with 0.018-inch or 0.025-inch coils. We typically use a 22-gauge needle for varices with a smaller diameter (< 10 mm) while a 19-gauge needle is used for varices with a larger diameter (> 10 mm). Once a needle is selected, remove the stylet and flush the needle with normal saline.

EUS-GUIDED EMBOLIZATION

Once a target is selected, the varix is punctured under endosonographic guidance. Coils are then deployed within the varix under endosonographic visualization using the stylet as a pusher. To prevent complications, keep the needle tip visible during coil deployment, avoid positioning against the counter wall of the varix, and flush the FNA needle with saline after each deployment to clear blood debris and prevent clot formation. Once an adequate number of coils appear to fill the varix, glue injection can be performed. Inject CYA glue in 0.5 cc aliquots until adequate flow obliteration is confirmed under Doppler examination. Allow 10 to 15 seconds between glue injections during examination to allow glue to solidify. It is important to flush the needle before withdrawal to expel residual glue which can damage the echoendoscope channel. Some recommend that glue should be flushed with distilled water as normal saline can induce premature solidification[32]. On average, for every 1 centimeter of varix diameter, we use 2 to 3 coils and 1 to 3 cc of glue. AGS may also be used instead of glue.

POST-EMBOLIZATION PHASE

Doppler ultrasonography is used post-procedure to confirm reduced blood flow within the treated varices and to detect potential residual variceal perfusion. Monitor for signs of early re-bleeding or any adverse effects. Although there is no standard interval for follow-up, we repeat EUS in 1 to 3 months, and 6 months thereafter to reconsider the need for re-intervention based on Doppler ultrasonography results.

INDICATIONS

Primary prophylaxis

Primary prophylaxis refers to the prevention of the first bleed. In high-risk patients, preventing GVH is of utmost importance given associated high transfusion requirements, uncontrolled bleeding, rebleeding, and death[33] (Figure 1). A few observational studies limited to a small number of patients reported promising results with EUS-coil for primary prophylaxis of GVs[30,34]. In a retrospective study, all 8 patients treated with EUS-coil for primary prophylaxis of GVH had no signs of bleeding nearly 1.5 years later[35]. Since, several groups reported additional benefit to combining coil and glue injection as this combination could more effectively embolize the GV while minimizing distal embolization risk. In a single-center observational study evaluating 80 patients with high-risk GVs (defined as > 10 mm in size, or cherry-red spots) for primary prophylaxis of GV bleeding, EUS-guided coil and CYA injection was highly effective at preventing gastric variceal bleeding[36]. In a large cohort of 28 patients who underwent EUS-coil/CYA injection for GV, 27 of them had complete obliteration of GV of EUS while the remaining patient had to be re-treated[31]. Similarly in another study, complete obliteration of varices was reported in all 6 patients who underwent primary prophylaxis with EUS-coil/CYA for GVH[37]. In a single center observational study, Kouanda et al[36] evaluated the efficacy of EUS-coil/CYA for primary prophylaxis of GV bleeding in 80 patients, and found that EUS-coil/CYA had a 96.7% rate of GV obliteration while only 2 patients experienced post-intervention bleeding, and adverse events were reported in 4 patients. A meta-analysis including 18 studies using EUS-coil, CYA, or a combination found that the pooled rate of GV obliteration in 162 patients undergoing primary prophylaxis was 90.2% with a re-bleeding rate of 4.9%[38]. For those using a combination EUS-coil/CYA therapy, the rate was 95.4%[38]. This highlights the efficacy of EUS-guided therapy for primary prophylaxis of GVH.

Figure 1 32-year-old male with history of cirrhosis secondary to alcohol use was referred for primary prophylaxis of gastric variceal bleeding. A: Upper endoscopy showed large type 1 isolated gastric variceal conglomerate; B: Endoscopic ultrasound showing a 3.7 cm variceal conglomerate; C: Trans-gastric deployment of coils through a 19-gauge needle; D: Endoscopic ultrasound showing thrombosis of targeted gastric varix; E: Follow-up endoscopy at 1 month showing smaller varices with cracked earth pattern suggesting obliteration; F: Absent flow on Doppler.

Despite the growing literature on the efficacy of EUS-coil for primary prophylaxis for GV hemorrhage, there is still limited long term data and absence of randomized clinical trials demonstrating the long-term benefit of these interventions[39]. Also, there is a lack of multicenter studies to guide which intervention (coils alone or with CYA) is most appropriate[40]. Consequently, the AASLD and the European Association for the Study of the Liver do not recommend EUS-coil for primary prophylaxis of GV hemorrhage[39].

Secondary prophylaxis and active bleeding

Secondary prophylaxis is defined as prevention of rebleeding (Figure 2). If left untreated, GVs will re-bleed in 50% to 80% of cases[41]. As such, preventing rebleeding is necessary as it is associated with a 20% risk of mortality[42]. Many studies have suggested the effectiveness of EUS- coil/CYA injection for that purpose compared to the conventional techniques[32]. Notably, Binmoeller[43] found that 23 out of 24 patients referred for secondary prophylaxis of bleeding GV had complete obliteration of varices after one session of EUS-coil/CYA. Rebleeding was observed in 1 patient. Similarly in another series, all 10 patients who underwent EUS-coil/CYA for secondary prophylaxis had complete obliteration of GV[37]. Several series have demonstrated the efficacy of EUS-guided coil/CYA for active bleeding. In one series, EUS-coil/CYA was performed on 7 patients with actively bleeding GV with effective hemostasis[19]. Several additional small series showed similar findings[29,44]. It is worth mentioning that a 2023 meta-analysis reported high treatment efficacy rates for EUS-guided interventions in secondary prophylaxis: 94.2% for EUS-glue, 95.5% for EUS-coil, and 88.7% for combination EUS-glue and coil therapies[38].

Figure 2 65-year-old female with history of metabolic dysfunction associated steatohepatitis cirrhosis presented with hematemesis and melena. A: Upper endoscopy showed large gastroesophageal varices type 2 with stigmata of recent bleeding; B: Endoscopic ultrasound (EUS) showing a large variceal conglomerate; C: The feeder vessel was located under EUS evaluation; D: Transesophageal deployment of multiple coils and absorbable gelatin sponge (Gelfoam slurry) under EUS guidance; E: Improvement in size of varices following EUS-guided coil embolization.

AGS is another synthetic tissue adhesive that has been combined with coil embolization for effective management of GV. Notably, gelatin sponges can absorb whole blood quantities up to 45 times their weight. In a series of 10 patients who underwent EUS-guided coil/AGS for secondary prophylaxis or actively bleeding GV, there was no reported rebleeding from GV after a mean follow-up of 6 months. Nine of the patients had follow-up EUS evaluation showing complete obliteration of GV[45]. While AGS is relatively novel with very limited data compared to glue embolization, this has been gaining traction over the past few years given its safety profile. To note, Gelfoam slurry needs to be manually prepared by mixing Gelfoam pad or powder with saline. In a multicenter study including 106 patients with GV, of which 71.1% presented with acute bleeding, EUS-coil with CYA or AGS had an 88.7% clinical success rate defined as cessation of bleeding or absence of bleeding at 30 days post-intervention[46]. Several other studies showed comparable results[32,47-50]. In a meta-analysis including 604 patients, the pooled rate for treatment efficacy for secondary prophylaxis of GV bleeding using EUS-coil/CYA was 88.7% and GV obliteration rate was 84.6% with a 20.6% risk of rebleeding[40]. This demonstrates that in expert hands, EUS-guided therapies are highly successful with a relatively low rate of rebleeding.

Other indications

Aside from GV treatment, EUS-coil is also effective for ectopic varices, vascular pseudoaneurysms, splenic artery embolization, and non-variceal bleeding[34,50,51]. For instance, Fujii-Lau et al[34] evaluated EUS-coil in 14 patients who failed standard ectopic varices therapy. They found that EUS-coils was successful for all patients. Ryou et al[50] demonstrated effective management of pseudoaneurysms with EUS-coils, especially when traditional radiologic interventions fail to achieve the desired results. Also, they pointed that EUS-guided embolization of the splenic artery can reduce splenic volume and lower portal pressure, thereby managing associated complications[50]. Finally, for non-variceal bleeding, EUS-guided embolization successfully terminated bleeding from Dieulafoy lesions and gastrointestinal stromal tumors, by directly targeting the bleeding vessel[51].

COMPARISON BETWEEN DIFFERENT ENDOSCOPIC INTERVENTIONS

Comparison of EUS-coil with direct endoscopic injection

Most of the literature involving EUS guided treatment of GV is limited to small case series and direct comparisons between different treatment methods are sparse. However, the literature clearly demonstrates that EUS-coil, especially when combined with CYA injection, has favorable outcomes compared to traditional endoscopic approaches. When compared to direct endoscopic injection of CYA, EUS-coil has been consistently associated with a lower rate of rebleeding and decreased risk of serious complications, such as systemic embolization[46-48]. Mukkada et al[52] compared 30 patients who underwent EUS-coil embolization with or without CYA with 50 patients treated with direct embolization with CYA for secondary prophylaxis of GV. Rebleeding rates were significantly higher in the direct endoscopic CYA injection group (51% vs 20%, P < 0.001). Similarly, Chen et al[48] compared the use of direct CYA injection to EUS-coil/CYA in patients with GV and spontaneous portosystemic shunts and found significantly lower rates of re-bleeding in the latter group (27.8% vs 4.8%, P = 0.041). In a randomized controlled trial comparing conventional CYA injection with EUS-coil/CYA in 32 patients with GV, asymptomatic pulmonary embolism was noted in a higher proportion of patients in the conventional CYA group, although the difference was not statistically significant (50% vs 25%, P = 0.14). A meta-analysis including 23 studies and 851 patients evaluated the efficacy, GV obliteration, rebleeding, and safety of EUS-guided therapy of GV compared to direct endoscopic injection. While there was no significant difference in treatment efficacy between the 2 groups, EUS-guided therapy had a significantly higher GV obliteration rate (84% vs 63%, P = 0.02), and lower rates of late rebleeding (12% vs 17%, P = 0.1) and GV recurrence (9% vs 18%, P = 0.06)[53].

Comparison between different EUS-guided interventions

Combining coils and CYA injection is more advantageous than CYA alone, as coils act as scaffold to CYA which enhance the obliteration of GV while decreasing the rates of systemic embolization. A randomized clinical trial compared EUS-coil/CYA and EUS-coils for the management of GV in 60 patients. Immediate GV obliteration was significantly higher in the combination group (86% vs 13.3%, P < 0.001) with lower rates of rebleeding (3.3% vs 20%, P = 0.04)[38]. A meta-analysis including 536 patients compared EUS-guided combination therapy of coil/CYA vs monotherapy. EUS-guided coil/CYA therapy achieved a significantly higher rate of gastric variceal obliteration compared to either EUS-CYA (98% vs 96%, P < 0.001) or EUS-coils alone (98% vs 90%, P < 0.001). It was also associated with a lower need for reintervention than both EUS-coils (15% vs 26%, P < 0.001) and EUS-CYA (15% vs 25%, P < 0.001). Moreover, the incidence of adverse events was significantly lower in the EUS-coil/CYA group compared with EUS-CYA (10% vs 21%, P < 0.001), though not significantly different from the EUS-coil group[54].

Financial implications

The financial implication of EUS-coil compared to conventional therapy should be considered. Romero-Castro et al[24] noted that 1 mL of CYA and one coil cost $72.30 and $99.4 respectively in the United States. Kouanda et al[36] claim that the total cost for EUS-coil/CYA is $1831 including a $1557 facility fee and $274 physician fee, while the cost of an inpatient hospitalization for GV bleeding reaches up to $11000.

COMPLICATIONS AND ADVERSE EVENTS

EUS-coil is associated with a low rate of overall adverse events estimated at 3% for EUS-coil and 10% for EUS-coil/CYA, which are significantly lower than EUS-CYA alone which has a 21% risk of adverse events[54]. Mild to moderate abdominal discomfort is relatively common post-procedure. A meta-analysis including 10 studies of patients with GV treated with EUS-coil/CYA, abdominal pain was reported in 9.79% of patients. This typically resolves within a few hours and is managed conservatively[47]. Minor self-limited intraprocedural bleeding may occur at the puncture site of the varix. Post-procedural bleeding is less common but may require additional intervention or blood transfusions. Bhat et al[31] reported a 5% incidence of intra-procedural bleeding and 3% incidence of post-procedural bleeding. Coils are occasionally seen extruding into the gastrointestinal lumen several weeks following coil embolization (Figure 3). This is a harmless complication and does not lead to increased risk of bleeding. Binmoeller[43] reported 3% of patients experienced coil extrusion. Coil migration has also been reported. Rowley et al[55] described one asymptomatic patient who experienced coil migration to the liver while another case reported the migration of a coil into the right ventricle. Tepox-Padron et al[56] described a patient who experienced hypoxia due to coil migration into the pulmonary artery. Potential reasons for coil migration include small coil size relative to the varix, deployment of the coil into the peri-GVs, or poor EUS visualization during coil deployment. Therefore, it is important to select an adequate coil size and accurately differentiate peri-gastric abdominal varices from submucosal varices. While risks of systemic embolization are lower with EUS-coil/CYA than CYA alone, there remains a 2% to 3% risk of distal embolization[43,46]. Fever and transient bacteremia can occur in up to 13% of patients[57,58]. Bick et al[28] report a patient who developed splenic infarcts and Escherichia coli bacteremia 2 days following the procedure requiring a prolonged hospital stay. The variation in reported adverse events between the studies is likely related to the study design and the significant variation in technique between studies including whether routine antibiotics were used for the procedure and amount of glue/coils administered. Mortality following the procedure has been reported but was largely attributed to unrelated causes[58].

Figure 3 24-year-old male with history of portal vein thrombosis and hepatic encephalopathy was recently hospitalized with hematemesis and found to have gastric varices on upper endoscopy. Imaging showed no evidence of gastro-renal shunt. He was referred for endoscopic ultrasound (EUS) guided therapy of gastric varices. A: Upper endoscopy showing isolated gastric varices type 1; B: EUS showing complete thrombosis of varices following EUS guided coil embolization and cyanoacrylate glue injection; C: Coil extrusion on follow-up endoscopy at 3 months; D: Follow-up endoscopy at 1 year showing complete eradication of varices.

CONCLUSION

EUS-coil is a promising technique for the management of GVs, providing a safer and more effective alternative to traditional approaches like CYA injection alone. This method allows improved visualization, enhanced efficacy, and reduced risk of significant complications including systemic embolization and re-bleeding[46-48]. Despite these benefits, there remain challenges in standardizing the technique and creating step-by-step guidance for its use, with limited high-quality data to guide the optimal approach. To date, there are no randomized controlled trials directly comparing NSBBs and EUS-guided therapy for primary prophylaxis, or TIPS/BRTO and EUS-guided therapy for secondary prophylaxis for GV bleeding. In addition, given the novelty of the procedure, there is limited long term data on the effectiveness and rebleeding rates of EUS-guided interventions. While the procedure is safe, complications such as abdominal pain, bleeding, infection, and systemic embolization can occur[6,31,43,56]. In addition, EUS-coil requires significant expertise that is not readily available at all centers. Further research, in particular large randomized clinical trials, are needed to continue to refine the technique, optimize patient selection, and define its role in primary and secondary prophylaxis of gastric variceal bleeding[39]. However, EUS-coil is a significant advancement in the management of GVs, offering the therapeutic endoscopists another valuable tool at their disposal.