NBCA is a liquid adhesive agent that can rapidly induce the process of polymerization and solidify when it comes in contact with solutions (such as blood) that contain anions, thereby promoting vessel obliteration, inflammation, and fibrosis[6]. Because of the high risk of pulmonary embolism, coils were used to slow down the blood flow during the embolization of varicose veins[7]. The migration of both the coil and NBCA is an extremely rare complication after TIPS, with only one case having been reported (Table 1). Kupkova et al[8]. incidentally found a metallic coil and NBCA that penetrated into the stomach three weeks after TIPS. Eleven months after TIPS, the coil spontaneously dropped from the gastric varices into the stomach without bleeding symptoms. In our case, the late-onset migration of the coil and NBCA occurred a year after TIPS, and the primary reason for migration in our patient may have differed from that of Kupkova et al. In our patient, the migration was most likely related to the recurrence of portal hypertension[9, 10]. The variceal lumen might have widened due to recurrent portal hypertension, leading to migration of the coil and NBCA. In addition, the immune response is a potential promigratory factor, as it functions to eliminate and isolate foreign material. Adhesion of macrophages at the surface of the foreign body and formation of a fibrous capsule with the release of degradation mediators are important components of the immunologic response[11]. Local inflammation and fistula formation may have also played a role[12].
Table 1
Summary of the cases of coil and NBCA migration following TIPS.
| Author/Date [Ref] | Age/Sex (y) | Previous liver disease | Vascular abnormality | Site of migration | Migration time from TIPS | Management | Outcome |
| Kupkova et al (2006) [8] | 68/F | Liver cirrhosis due to hereditary haemochromatosis | Gastric varices | Stomach fundus | 3 weeks | None | Well |
| Present case | 46/M | Hepatitis B cirrhosis | Gastroesophageal varices | Stomach fundus | 1 year | None | Well |
| TIPS, transjugular intrahepatic portosystemic shunt; NBCA, n-butyl-2-cyanoacrylate |
Separate coil migration and penetration following TIPS placement are also rarely documented, with only seven cases having been reported in the English literature (Table 2)[9, 10, 12–16]. Because of the rarity of such complications in the literature, it is difficult to make recommendations for the management of this condition. However, considering the different clinical evolutions of such cases, each patient requires an individual assessment. Three case reports have described coil migration without active bleeding following TIPS creation. No further intervention was performed for the migrated coils[12–14]. Merchant et al[9]. reported a case of coil erosion through a gastric varix into the gastric lumen without active hemorrhage after liver transplantation. The migrated coil was left in situ. Unfortunately, the patient died of progressive dyspnea and subsequently developed polymicrobial sepsis. Similarly, another case reported coil penetration into the stomach after liver transplantation. Urgent laparotomy and partial gastrectomy were performed because of massive hematemesis. Unfortunately, the patient died two days after the surgery[15]. Lucas et al[10]. described a case of gastrointestinal hemorrhage due to coil extrusion after TIPS creation in a patient treated with sorafenib. The portal stent was repermeabilized, and the gastric varices were embolized using a plug. Another report described embolization coil erosion through the duodenal varix following TIPS creation. The penetrating coil was removed using endoscopic forceps[16]. In our case, the patient refused further intervention to retrieve the coil and instead received conservative therapy. Appropriate treatment should be given once signs of bleeding are noted to ensure a good prognosis for the patient. Therefore, close surveillance with endoscopy is recommended for detecting the coil and varices.
Table 2
Summary of the cases of separate coil migration following TIPS.
| Author/Date [Ref] | Age/Sex (y) | Previous liver disease | Vascular abnormality | Number of coils insertion | Site of coil migration | Time from coil insertion | Management | Outcome |
| Lucas et al (2011) [10] | 49/F | Alcoholic and hepatitis C cirrhosis | Gastric varices | 2 | Stomach fundus | 2 years | Portal stent was repermeabilized and the gastric varice was embolized using a plug | Well |
| Hussain et al (2011) [12] | 62/M | Hepatitis C cirrhosis | Gastric varices | 3 | Stomach fundus | 2 weeks | None | Well |
| Merchant et al (2013) [9] | 55/F | Hepatitis C cirrhosis | Gastric varices | 4 | Stomach fundus | Approximately 3 years | None | Died |
| Oza et al (2013) [13] | 61/F | Hepatitis C cirrhosis | Duodenal varices | 14 | Duodenal lumen | 3 months | None | Well |
| Levi et al (2013) [15] | 64/F | Hepatitis C cirrhosis | Gastric varices | Unknown | Stomach fundus | Approximately 3 years | Surgical removal (partial gastrectomy) | Died |
| Soape et al (2017) [16] | 65/F | Primary biliary cholangitis and decompensated cirrhosis | Duodenal varices | Unknown | Cecum | 1 month | Endoscopic removal | Well |
| Pusateri et al (2020) [14] | 58/M | Alcoholic cirrhosis | Gastric varices | Unknown | Stomach fundus | Approximately 2 years | None | Well |
| TIPS, transjugular intrahepatic portosystemic shunt |
There are currently no recommended definitive treatments for such cases. However, from our experience, we may be able to provide information for future cases, as this complication should be based on individual patient assessments, such as in our case. Although the outcome of this complication is unknown to date, it may potentially contribute to rebleeding or gastric perforation[14]. Furthermore, there is a likelihood of spontaneous dropout of the coil, similar to that of Kupkova et al. Because the migration of the coil and NBCA in our patient was likely related to recurrent portal hypertension, long-term follow-up is mandatory not only to detect such late complications as soon as possible but also to evaluate the progression of cirrhosis, stent patency, and portal thrombosis. Nonetheless, when such complications occur, it is essential to consider the case thoroughly, react appropriately, and learn from the experience[17].
The current case presents an extremely rare but significant complication after TIPS. Our report highlights the management and follow-up recommendation for such rare cases. Since this is only the second case of simultaneous migration of a coil and NBCA, our experience may provide guidance for the management of future similar cases and stimulate discussion about treatment methods of similar patients.