The role of partial splenic artery embolization in the management of refractory esophageal variceal bleeding due to portal vein thrombosis

Abstract

Background

Gastro-esophageal variceal hemorrhage (GEVH) is one of the major causes of life-threatening gastrointestinal bleeding in children. Medical, endoscopic, angiographic, and surgical interventions can be utilized in treatment. In this case report, we describe partial splenic artery embolization for refractory GEVH due to portal vein thrombosis.

Case presentation

A 3-year-old male patient was admitted with abdominal distension. He had been followed up for hepatosplenomegaly for 2 years. The patient's body weight was 15.5 kg (50-75th percentile, 0.69 SDS) and height was 96 cm (50-75th percentile, 0.27 SDS). The general condition was moderate, and the skin appeared pale. The liver was palpable 2 cm, and the spleen was palpable 6 cm below the costal margin. Other system examinations were normal. Laboratory findings included hemoglobin (Hgb) of 7.1 g/dL, hematocrit (Hct) of 24%, white blood cell count of 9800/mm3, platelets of 67000/mm3, and INR of 1.3. Abdominal CT angiography demonstrated a hypodense thrombus at the portal confluence, almost completely occluding the lumen. Endoscopy revealed esophageal varices at the 2 and 7 o'clock positions in the distal esophagus, which became more prominent with insufflation and had red spots on them. Hyperemia was observed in the corpus and antrum of the stomach. Findings were consistent with stage 3 esophageal varices and increased vascularity in the duodenum due to portal vein thrombosis. Medical treatment with a proton pump inhibitor and beta-blocker was initiated. The patient underwent endoscopic band ligation (EBL) three times over 9 months. Despite EBL, the patient presented with GEVH three times during a 1.5-year follow-up. Due to newly developed multiple varices observed on control endoscopy a decision was made to perform splenic artery embolization. Interventional radiology performed selective lower splenic pole embolization. Six months later, the patient underwent another selective embolization. The patient has been followed up for 3 years without GEVH.

Conclusions

In this case, splenic artery embolization was observed to be an effective, repeatable, and safe treatment method for patients with hypersplenism caused by portal hypertension and refractory esophageal variceal bleeding.

Background

Gastroesophageal variceal hemorrhage (GEVH) is a serious and life-threatening complication in patients with portal hypertension (PH) [1–3]. Although overall mortality has decreased in the last decade, recurrence rates of GEVH can reach 70%, and approximately 20–50% of cases result in mortality [2]. Pharmacological, endoscopic, angiographic, and surgical interventions are the four main treatment strategies for GEVH [3–6]. These strategies are designed to achieve three therapeutic goals by reducing portal system pressure: primary prophylaxis, treatment of acute variceal bleeding, and secondary prophylaxis [1, 3, 5]. Each treatment approach has its own advantages and disadvantages. Typically, patients with mild to moderate GEVH are treated with endoscopic interventions such as endoscopic band ligation (EBL), endoscopic injection sclerotherapy (EIS), and pharmacological treatments. Combinations of pharmacological and endoscopic treatments can be used for acute bleeding episodes and secondary prophylaxis of variceal bleeding [7, 8]. Splenic artery embolization is an interventional radiological method that aims to reduce portal system pressure in stable patients. It has been suggested that splenic artery embolization is an effective method in cases of resistant variceal bleeding [6]. In this case report, the application of partial splenic artery embolization to a patient who developed resistant esophageal variceal bleeding secondary to portal vein thrombosis due to umbilical vein catheterization will be described.

Case presentatıon

A 3-year-old male patient presented to the gastroenterology outpatient clinic with restlessness and abdominal distension. His medical history revealed a two-year follow-up for hepatosplenomegaly. He was born prematurely as a twin at 32 weeks via cesarean section, and received umbilical vein catheterization (UVC) in the neonatal intensive care unit (NICU). There was no consanguinity or significant family history.

On physical examination, his weight was 15.5 kg (50-75th percentile, 0.69 SDS), height was 96 cm (50-75th percentile, 0.27 SDS), and he appeared fair with pale skin. Cardiovascular and respiratory systems were normal. The liver was palpable 2 cm, and the spleen was palpable 6 cm below the costal margin. Laboratory tests showed Hgb: 7.1 g/dl, Hct: 24%, WBC: 9800/mm3, platelets: 67000/mm3, and INR: 1.3 (Table 1).

Table 1.

Laboratory parameters

	Before Embolization	After Embolization
Hb (g/dL)/Hct (%)	7.1 / 24	10.1/31
WBC (count/mm3)	9800	10230
Platelets (count/mm3)	67000	240000
AST (U/L)	31	34
ALT (U/L)	20	25
ALP (U/L)	208	231
GGT (U/L)	15	17
LDH (U/L)	320	340
Total / Direct Bilirubin (mg/dL)	0.9 / 0.2	1/0.3
Albumin (gr/dL)	2.8	3.4
Ammonia (mmol/L)	35	30
AFP (ng/mL)	4.5	4.2
Acid Phosphatase (U/L)	5.1	5
Prothrombin Time (sec)	22	20
INR	1.3	1.2

Hb Hemoglobin, Hct Hematocrit, WBC White blood cell, AST Aspartate transaminase, ALT Alanine transaminase, ALP Alkaline phosphatase, GGT Gama-glutamyl transferase, LDH Lactate dehydrogenase, AFP Alpha-fetoprotein

Abdominal ultrasound revealed a grossly heterogeneous liver, consistent with diffuse liver disease. Abdominal CT angiography showed a thrombus extending from the portal vein confluence to the SMV, almost completely occluding the lumen, along with associated collaterals and splenomegaly (Fig. 1).

Fig. 1.

In the portal venous phase (45th second). a A hypodense filling defect consistent with a thrombus is observed, nearly occluding the lumen in the distal section of the portal vein confluence (orange circle). b A hypodense filling defect consistent with a thrombus is seen, almost completely occupying the lumen in the distal section of the portal vein confluence (orange ellipse). c A filling defect consistent with a chronic partial thrombus, accompanied by luminal narrowing, is identified in the pre-confluence segment of the SMV (red square)

During endoscopy, esophageal varices with red spots were noted at 2 and 7 o'clock, which did not blanch with insufflation and were more prominent in the lower esophageal segment. Gastric corpus and antrum showed signs of hyperemia. Findings were consistent with stage 3 esophageal varices and increased duodenal vascularity.

The patient received proton pump inhibitor and beta-blocker treatment. EBL was performed thrice over 9 months, yet despite treatment, he experienced recurrent GEVH over 1.5 years. Subsequent computerized tomography angiography revealed newly developed multiple varices, prompting a decision for splenic artery embolization. Interventional radiology achieved embolization of the splenic artery branch to the lower pole using coils and 700–900 micron particles (Fig. 2).

Fig. 2.

Angiographic images. a Angiographic image of the splenic artery and spleen before embolization, demonstrating splenomegaly. b Following the first embolization, coil embolization was performed on the inferior polar branch of the splenic artery, resulting in a 25% reduction in parenchymal staining. c During the second embolization session, coil embolization was performed on the middle polar branch of the splenic artery, followed by particle embolization. Compared to the first embolization, splenic parenchymal staining decreased by approximately 50%

The patient remained stable post-procedure, with uncomplicated splenic infarcts noted on ultrasound. Follow-up was uneventful, and he was discharged with medical management. Approximately two months later, he returned with melena, prompting suspicion of recurrent GEVH. Repeat partial splenic artery embolization targeted the segmental artery feeding the middle lobe of the spleen. Following the procedure, the spleen reduced to 4 cm, platelet count stabilized at 240,000/mm3, and the patient remained free of GEVH during three years of follow-up.

Discussion and conclusions

Umbilical venous catheterization (UVC) is a commonly utilized vascular access method in neonatal intensive care units (NICU) and is associated with potential serious side effects [9]. There are case reports of portal vein thrombosis (PVT) induced by UVC reported in the literature [10–12]. The technique of catheter insertion and the duration of catheter use have been observed to impact complications in these patients [13, 14]. In their study evaluating PVTs developing in the neonatal period, Kim JH et al. [14] reported that 43% of the cases had asymptomatic PVTs, and 56% of neonatal PVTs followed showed spontaneous complete or partial resolution. It was observed that thrombus size was smaller in cases with spontaneous regression, while catheterization and blood transfusion for longer than 6 days increased the risk of PVT.

In the literature, asymptomatic clinical courses have been reported in children who developed PVT due to UVC during NICU admission, presenting with GEVH around the age of 3–5 years [11, 12]. Medical, endoscopic, angiographic, and surgical methods can be employed to manage GEVH in these patients. This patient group generally has a more favorable prognosis compared to GEVH caused by cirrhosis. In their series examining 63 pediatric patients with extrahepatic PVT, Grama A et al. [15] reported mortality due to late complications and severe sepsis in only one case (1.6%). In cases of GEVH that cannot be controlled by medical and endoscopic interventions, angiographic methods or, as a last resort, surgical interventions can be considered. Surgical intervention is typically indicated for unstable massive GEVH [15]. Angiographic interventions can effectively lower portal vein pressure and mitigate the risk of GEVH in stable patients, with splenic artery embolization being a recognized method for this purpose. The spleen plays a crucial role in immune function and its removal, such as through asplenia, especially in the first 5 years of life, can lead to severe sepsis. Therefore, in splenic artery embolization aimed at reducing portal vein pressure, selective embolization of one of the polar branches of the splenic artery is typically performed [6]. In the patient described in this case report, the splenic artery was selectively catheterized, and in the first session, the lower polar branch was embolized. Subsequently, due to persistent GEVH, the middle polar branch was embolized in the second session.

Early clinical follow-up is crucial to detect complications following splenic artery embolization. Post-embolization syndrome occurs in approximately 75% of cases within the first 24–72 h and is typically self-limiting. Rarely, complications such as pancreatitis, left-sided pleural effusion, splenic abscess, and splenic rupture may occur. The mortality rate associated with splenic artery embolization has been reported as 0.5–1% in the literature [16]. No complications have been observed in our case, which continues to be followed-up.

In conclusion, it can be said that splenic artery embolization may represent an effective and reliable treatment alternative for patients with hypersplenism due to PH and resistant GEVH.

Abbreviations

EBL

Endoscopic band ligation

EIS

Endoscopic injection sclerotherapy

GEVH

Gastro-esophageal variceal hemorrhage

INR

International normalized ratio

NICU

Neonatal intensive care unit

PH

Portal hypertension

PVT

Portal vein thrombosis

UVC

Umbilical venous catheterization

Authors’ contributions

Conceptualization – BTD, NU, AHB, ÖD, İÖ; Data curation – BTD, NU, AHB, ÖD, İÖ; Investigation – BTD, NU, AHB, ÖD, İÖ; Methodology – BTD, NU, AHB; Project administration – NU, AHB, ÖD, İÖ; Supervision – NU, AHB, ÖD, İÖ; Validation – AHB, ÖD, İÖ; Visualization – BTD, NU, AHB, ÖD, İÖ; Writing – original draft – BTD, NU; Writing – review & editing – NU, AHB, ÖD, İÖ All authors read and approve final version of manuscript.

Funding

Not applicable.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Written informed consent waa taken from patient’s legal guardians for publication of this case report.

Consent for publication

Written informed consent was taken from patient’s legal guardians for publication of this case report and for use of radiologic images.

Competing interests

The authors declare no competing interests.