Endoscopic ultrasound–guided coil embolization and absorbable hemostatic gelatin sponge is a safe and efficacious treatment for bleeding rectal varices

Abstract

Background and Aims

Rectal variceal bleeding is a rare life-threatening cause of lower gastrointestinal bleeding. We present the first case series to our knowledge of successful hemostasis achieved with combination of endoscopic ultrasound (EUS)-guided coil embolization and absorbable gelatin sponge (AGS) for bleeding rectal varices.

Methods

We performed a retrospective review of 6 patients with active rectal variceal bleeding. All patients underwent EUS-guided coil embolization with AGS therapy. Procedural details, technical success, adverse events, and rebleeding rates were measured.

Results

We achieved 100% technical success (6 of 6 patients) with successful sonographic rectal varices decompression. A range of 1 to 4 coils with a total length of 7 to 20 cm was injected along with AGS. There were no adverse events or rebleeding from recurrent rectal varices. Median follow-up period was 190 days.

Conclusions

EUS-guided coiling coupled with AGS therapy appears to be a potentially safe and efficacious alternative for patients with bleeding rectal varices.

Introduction

Rectal varices (RVs) are defined as the abnormally dilated collateral vessels that occur secondary to systemic portal hypertension.¹ RVs are dilated circumferential submucosal veins that are seen 4 to 7 cm above the anal verge, noncontiguous with anal columns, and distinct from internal hemorrhoids. They serve as collaterals between superior rectal vein of the inferior mesenteric system (portal circulation) and the middle and inferior rectal vein of the iliac system (inferior vena caval circulation). Various shunts include inter-rectal and rectogenital shunts that provide venous communication hubs between rectal venous plexus.²

Rectal variceal bleeding is a rare cause of lower gastrointestinal bleeding that accounts for 1% to 5% of all variceal bleeding.³ Band ligation and cyanoacrylate glue injection are potential endoscopic therapies but have been associated with high rates of recurrence of rectal variceal bleeding because of rich collateral blood supply as compared with esophageal varices.⁴ Interventional radiology—guided therapies such as a transjugular intrahepatic portosystemic shunt (TIPS) can be considered in these patients to decrease the overall portal pressure to prevent rectal variceal bleeding, but may be contraindicated in patients with high Model for End-Stage Liver Disease-Sodium (MELD-Na) scores, portal vein thrombosis, and significant cardiopulmonary comorbidities.⁴ Surgical shunting, including portocaval, mesocaval, and splenorenal shunts, has been associated with morbidity in patients with decompensated cirrhosis and often is avoided because of the risk of worsening hepatic encephalopathy. Under-running suturing or direct ligation of varix often is temporary, and recurrence bleeding is common.¹

Lower endoscopic ultrasound (EUS) is currently the most sensitive test for diagnosing RVs, which appear as longitudinal, submucosal tortuous structures within the rectal wall.⁵ Given the ease of visualization of RVs with EUS and lack of a standardized treatment, therapeutic EUS modalities should be considered as viable alternatives for targeted rectal variceal therapy. EUS-guided coiling therapy specifically in combination with glue is a promising endoscopic therapy that has demonstrated remarkable success in the treatment of gastric varices. Several studies have shown excellent clinical and technical success for this technique over direct endoscopic glue injection.⁶ Therefore, we applied the EUS-guided coiling embolization technique toward treatment of RVs, which we previously described as a video case report (included in this case series), specifically in targeting of the perforator vein that primarily fed the various variceal nests.⁷ In addition, we applied absorbable gelatin sponge (AGS) as an adjunct therapy to complement hemostasis effects. AGS is a hemostatic agent made of purified collagen that has been historically used by various surgical subspecialties and interventional radiology to control intraoperative bleeding.⁸

To date, evidence on EUS-guided coil embolization for rectal variceal bleeding remains limited to individual case reports. To our knowledge, we present the first institutional case series describing the use of adjunctive AGS therapy in conjunction with EUS-guided coiling for the treatment of bleeding RVs in 6 hospitalized patients.

Methods/case description

We conducted a retrospective review of prospectively collected data on patients with active lower gastrointestinal bleed secondary to RVs who had undergone EUS-guided coil embolization and AGS between October 2018 and March 2024 at a single tertiary care center. All patients were referred for EUS-guided procedures for rectal variceal treatment after multidisciplinary review.

The baseline demographics and procedural characteristics were obtained for each patient from the hospital electronic medical record system. The procedure details were obtained from endoscopy reports and archived videos. Patient data extracted included laboratory testing and associated serum-based liver disease severity scores (eg, MELD-Na, Child-Pugh score) as well as procedure-related findings, such as maximal cross-sectional diameter of RVs (in millimeters), number and length of coils injected, and postprocedural adverse events.

Procedure details

All procedures were performed with the patient under general anesthesia and in the left lateral position. All patients underwent initial white-light colonoscopy to determine the cause of hematochezia and subsequently to identify presence of RVs. The colonoscope was exchanged for a linear echoendoscope (EG-580UT; Fujifilm Healthcare Americas, Lexington, Mass, USA) to confirm the diagnosis of RVs and mark the rectal variceal anatomy and presence of shunts. The rectal lumen was then filled with saline for optimal visualization of rectal variceal nests and associated shunts. On EUS, RVs were characterized as multiple circumferential anechoic tubular structures. EUS-guided coil injection was performed with preferential targeting of the perforator vein, which is responsible for feeding the variceal nests. The EUS needle was advanced to the most distal target of the variceal nest, which was then injected with one or multiple coils until the variceal compartment was sufficiently packed with coils. The needle was then withdrawn and repositioned to target additional compartments for further coil placement, if available. Choice of coil size (Nester Embolization Coils, Cook Medical, Bloomington, Ind, USA) include 0.035 inches (compatible with the 19-gauge needle) or 0.018 inches wire diameter (compatible with the 22-gauge needle) with cumulative coil lengths ranging from 7 to 20 cm. After coil injection, AGS slurry (Pfizer, New York, NY, USA) or SURGIFLO liquid hemostatic matrix (Johnson & Johnson Wound Management, Somerville, NJ, USA) was then injected for hemostatic reinforcement (1-2 mL), using the prior coils as a scaffold for localized polymerization. To create the slurry, we cut AGS strips into small pieces and packed them into a 10-mL syringe to create the slurry. A separate 10-mL syringe was filled with 5 mL of saline, and the 2 syringes were connected via a 3-way stopcock and mixed together until the end product was a liquid slurry.⁹ All patients received prophylactic antibiotics before the procedure to minimize the risk of procedure-related infection. Technical success was defined as successful injection of coils followed by AGS into the selected rectal varix leading to immediate reduction or cessation of blood flow as confirmed by Doppler. Successful rectal variceal decompression was defined as greater than 75% size reduction endoscopically and/or greater than 75% Doppler flow reduction sonographically.

Results

Six patients underwent EUS-guided coil embolization for treatment of RVs (a representative patient with images and video of endoscopic and EUS views are provided in Fig. 1, Video 1, available online at www.igiejournal.org). Preprocedural demographic and patient characteristics are summarized in Table 1. Mean age was 62 years; 3 (50%) patients were female. Five of 6 cases presented with hematochezia only. One case had both hematochezia and melena. Regarding the etiology of portal hypertension, 4 (67%) cases had a cirrhosis-related cause of portal hypertension. Mean Child-Pugh score was class B with mean MELD-Na of 13. Three patients required blood preprocedural transfusions. All patients were considered to have refractory RVs because of previous unsuccessful endoscopic therapy or ongoing bleeding. Surgical and endovascular options (eg, TIPS) were deferred because of high MELD-Na scores, portal vein thrombosis, or significant comorbidities (eg, hepatic encephalopathy).

Figure 1.

Representative endoscopic and endoscopic ultrasound (EUS) views of rectal varices before and after treatment. A, Endoscopic image showing rectal varices (red arrows) and internal hemorrhoids (blue arrows). B, EUS view of the rectal variceal nests with diameter measurements of the variceal vessels (bottom left corner, with 3.0 mm for yellow line and 5.0 mm for green line). C, EUS-guided fine-needle aspiration (EUS-FNA) injection of coils (blue arrows pointing to EUS-FNA needle, red arrows pointing to coils). D, EUS view demonstrating reduction in Doppler flow after treatment.

Table 1.

Preprocedural demographic characteristics of patients with rectal varices

Case ID	Age, years	Sex	Etiology of portal hypertension	Bleeding presentation	Child-Pugh classification	MELD-Na score	Hemoglobin, g/dL	Platelets, k/μL	INR	Preprocedural no. units of pRBCs transfused
1	79	Female	MASH cirrhosis	Hematochezia	A	10	10.7	96	1.2	0
2	59	Female	Oxaliplatin-induced liver injury	Hematochezia	B	9	7.2	114	1.3	3
3	59	Female	Nodular regenerative hyperplasia	Hematochezia	A	11	8.4	159	1.1	0
4	71	Male	Alcoholic cirrhosis	Melena and hematochezia	B	12	8	78	1.7	0
5	74	Male	Cirrhosis from AIH/PBC	Hematochezia	B	14	6	69	1.3	2
6	31	Male	Alcoholic cirrhosis	Hematochezia	C	24	7.4	34	2.1	3

AIH, Autoimmune hepatitis; ID, identification; INR, international normalized ratio; MASH, metabolic dysfunction-associated steatohepatitis; MELD-Na, Model for End-Stage Liver Disease–Sodium; PBC, primary biliary cholangitis; pRBC, packed red blood cell.

Procedural characteristics are summarized in Table 2. Technical success was seen in all cases (100%). All 6 patients showed rectal variceal decompression, with reduced submucosal prominence on endoscopy and decreased Doppler flow on EUS. A range of 1 to 4 coils with a cumulative length of 7 to 20 cm was injected. AGS was used in 100% of cases. All cases (100%) had concomitant hemorrhoids. No cases resulted in intraprocedural or immediate postprocedural adverse events (including severe bleeding or new hemodynamic instability). None required postprocedural blood transfusion. No patients had recurrent rectal variceal bleeding. Median follow-up period was 190 days.

Table 2.

Procedural characteristics of patients with rectal varices

Case ID	Maximum cross-sectional diameter of varices, mm	No. of coils injected	Coil specs	Technical success	Adverse events	Follow-up time, d	Rebleed from rectal varices
1	3.0	4	Four 0.018-in × 7-cm	Y	N	1460	N
2	2.5	1	One 0.018-in × 7-cm	Y	N	475	N
3	4.0	2	Two 0.018-in × 7-cm	Y	N	28	N
4	5.0	2	One 0.035-in × 20-cm One 0.018-in × 7-cm	Y	N	58	N
5	7.0	2	Two 0.018-in × 7-cm	Y	N	200	N
6	3.0	1	One 0.018-in × 7-cm	Y	N	180	N

ID, Identification; N, no; No., number; specs, specifications; Y, yes.

Discussion

We present a case series of hospitalized patients with refractory rectal variceal bleeding who were deemed unsuitable candidates for surgery or endovascular therapies, ultimately undergoing successful EUS-guided coil-based injection therapy. We demonstrated 100% clinical and technical success with EUS-guided coil embolization and adjunct AGS in bleeding RVs. Per review of the literature, this is the first reported case series to our knowledge on combination of EUS-guided coil and AGS injection for treatment of RVs. All patients demonstrated both endoscopic and sonographic evidence of decompression of RVs with Doppler flow reduction and no evidence of postprocedural adverse events.

There is a longstanding precedence in the interventional radiology literature on use of coil injection therapy and AGS for treatment of bleeding varices, often as an adjunct to TIPS and with an excellent safety profile.¹⁰ An EUS approach provides the added advantage of more-targeted therapy of the perforator vein that feeds the variceal nests, which can be difficult to achieve with interventional radiology–guided endovascular therapy. An adjunct therapy, such as AGS or cyanoacrylate glue, can then be subsequently injected at the coil injection site, using the coil as a scaffold to further complement hemostasis. Overall, EUS-guided coil embolization and AGS injection have minimal risk of adverse events, which may include bleeding, infection, and systemic embolization.¹¹ We prophylactically gave patients antibiotics periprocedure to reduce the risk of infection, although there are few data to support this practice. For risk of systemic embolization, at our institution we prefer AGS as the adjunct therapy, given the relatively lower theoretical adverse event profile compared with cyanoacrylate glue, which has demonstrated rare cases of systemic embolization in the gastric varices literature.

Although our case series presents encouraging results, there are several limitations. First, this is a single-center, single-operator study involving a small number of patients with limited short-term follow-up, which severely limits the scope of the study and potential conclusions. Second, the study is retrospective in design and is prone to selection bias. Lastly, given the referral pattern at our institution for EUS-guided coiling of RVs (ie, patients who were deemed not suitable for surgery or endovascular therapies), we were not able to find a suitable comparison arm for comparative analysis.

In conclusion, EUS-guided coil embolization with AGS is a viable endoscopic treatment option for bleeding RVs. We demonstrated good hemostasis outcomes and a low adverse event profile with our limited case series. Follow-up prospective cohort studies, possibly multicenter in design, will help further elucidate the role and utility of EUS-guided coil embolization with AGS for treatment of RVs.

Patient Consent

All participants provided consent for the procedure and study was approved by our hospital institution review board.

Disclosure

The following authors disclosed financial relationships: M. Ryou: Consultant for Olympus and Fujifilm. All other authors disclosed no financial relationships.

Supplementary data

Video 1

Video showcasing EUS views of a patient case who underwent EUS-guided coil embolization therapy with absorbable gelatin sponge for treatment of rectal varices.