Gastric cancer represents the fifth most common malignancy that affects humans and accounts for 8.8% of worldwide cancer-related deaths annually.¹ Curative disease is defined as gastric cancer without distant metastasis (stage 0–III). Radical surgical resection, accomplished by total or distal gastrectomy, represents the mainstay of curative treatment for gastric cancer; however, recurrent cancer occurs in up to 50% of patients, mostly within 1 year, and in individuals with more advanced disease.² Recurrent cancer is rarely curable because it is usually presents with diverse forms and at multiple sites.³ Patients with recurrent cancer often experience debilitating symptoms, such as nausea, vomiting, dysphagia, dehydration, and malnutrition, because of malignant gastric-outlet, duodenal, and jejunal obstructions, which can result in progressive deterioration of the patient’s quality of life.⁴ However, palliative bypass surgery is usually not amenable to patients with recurrent cancer and is associated with a high risk of both morbidity and mortality.⁵

Self-expandable metal stent (SEMS) placement is a well-established minimally invasive alternative to surgery for the palliative treatment of a malignant gastric outlet obstruction.⁶ This approach has been shown to be more advantageous than surgical bypass as it minimizes the time to re-establish oral intake and the length of hospital stay, although it is associated with higher rates of complication and reintervention and a shorter patency duration as a consequence of stent malfunction (i.e., because of tumor in-/over-growth and stent migration).^7–10 However, SEMS placement in patients with recurrent cancer is generally considered to be more technically difficult and complicated because of the anatomical alterations induced by radical gastrectomy. This review discusses the anatomical and technical considerations along with the outcomes of SEMS placement for the palliative treatment of recurrent cancer in patients with the surgically-altered stomach. Additionally, potential future advances in SEMS placement for patients with recurrent cancer are also discussed.

Total gastrectomy

Total gastrectomy refers to gastric resections that remove the entire stomach. This operation is mostly used for gastric cancers in the proximal (upper third) part of the stomach.¹¹ Cases of proximal gastric cancer that do not involve the esophagogastric junction can be approached by either total or proximal gastrectomy. However, most surgeons prefer the former because it is associated with a lower incidence of reflux esophagitis after gastrointestinal reconstruction.¹² After total gastrectomy, the gastrointestinal tract requires restoration of the enteric flow between the esophagus and small intestine. More than 60 methods for reconstruction have been described since the first successful total gastrectomy was reported in 1897; however, no consensus exists regarding an optimal approach.¹³ Most commonly, gastrointestinal continuity is restored by a Roux-en-Y reconstruction: this involves division of the upper jejunum and fashioning of an anastomosis between the remnant esophagus and the distal jejunal limb, with or without a pouch. This is followed by an anastomosis of the proximal jejunal limb draining the liver (the ROUX loop) to the distal jejunum approximately 20 to 40 cm below the esophagojejunosotmy (jejunojejunostomy; Fig. 1).¹⁴

Distal gastrectomy

Distal gastrectomy encompasses all types of gastric resection that do not involve the esophagogastric junction (i.e., antrectomy as well as two-thirds and four-fifths gastrectomy). This operation is generally used to treat gastric cancers located in the distal (lower two-thirds) part of the stomach.¹¹ After distal gastrectomy, restoration of the enteric flow between the remnant stomach and small intestine is required to restore the gastrointestinal tract. However, selecting an appropriate method to restore gastrointestinal continuity is controversial and has been left to the discretion of the surgeon¹⁴; the most frequently used restoration methods are Billroth I, Billroth II, and Roux-en-Y reconstructions.

Billroth I reconstruction is characterized by an anastomosis between the remnant stomach and duodenal stump (gastroduodenostomy; Fig. 2A). This method of reconstruction is generally restricted to cases with an antrectomy carried out because of the limited mobilization capabilities of the stomach and duodenum to create tension-free anastomosis. Although the Billroth I reconstruction method is infrequently performed in Western countries, it is commonly used in Asia.

Billroth II reconstruction is characterized by anastomosis between the remnant stomach and first jejunal loop (gastrojejunostomy; Fig. 2B). Additionally, a side-to-side jejunojejunostomy between the afferent and efferent loops can be established to reduce postoperative complications, such as afferent loop syndrome and delayed gastric emptying (Fig. 2C). However, most surgeons prefer not to perform a jejunojejunostomy because of the potential risk of anastomotic complications. Billroth II reconstruction is traditionally used when a Billroth I reconstruction is not feasible.

Roux-en-Y reconstruction is characterized by anastomosis between the remnant stomach and a distal jejunal limb (gastrojejunostomy) as well as an end-to-side anastomosis between the the first jejunal loop and distal jejunum (jejunojejunostomy; Fig. 2D). This method of reconstruction is an alternative to Billroth II reconstruction, and is more commonly used in Western countries than in Asia.

Currently available SEMS for the upper gastrointestinal tract include those used for esophageal and gastroduodenal obstruction.¹⁵ These devices are either woven, knitted, or laser-cut from stainless steel or alloys (i.e., elgiloy and nitinol) into a cylindrical shape, and differ in structural and mechanical properties. Some SEMSs are fully or partially-covered with a covering membrane (i.e., silicone, polyurethane, polytetrafluoroethylene, or nylon) to prevent tumor ingrowth, and some are flared at either the proximal or both ends to reduce stent migration.¹⁶ An esophageal SEMS is typically a fully-covered stent that has a thread attached to the proximal end so that it can easily be retrieved, whereas a gastroduodenal SEMS is generally a bare or partially-covered stent that is not designed to be removable.¹⁵ For insertion, the SEMS is compressed and loaded into the delivery system and then deployed through the working channel of the endoscope or over a guide wire under fluoroscopy.

The optimal selection of a SEMS for malignant gastroduodenal obstruction is controversial, as no consensus of opinion exists for any particular stent.^15,16 Nevertheless, partially-covered SEMSs are being used increasingly because of their resistance to tumor ingrowth and acceptable low migration rates.^17–19 Notably, a recent multicenter randomized control trial has shown significantly higher long-term patency with partially-covered SEMSs compared with bare SEMSs.²⁰ Although this study mainly included patients without previous gastric surgery, the findings should be generally applicable to those with recurrent cancer after radical gastrectomy.

Peroral self-expandable metal stent placement

SEMSs are typically placed via the peroral route under endoscopic and/or fluoroscopic guidance. It is generally thought that the outcomes of this procedure are similar irrespective of whether the stent is placed under endoscopic and/or fluoroscopic guidance.^6,21 The authors of this present study are radiologists who typically perform SEMS placement under fluoroscopic guidance alone (Fig. 3). However, a combined approach is particularly useful, when the likelihood of passing the stricture with catheter and guide wire alone is low, such as with a long or tortuous access route.¹⁷ Fluoroscopic SEMS placement can be performed under local anesthesia alone, but sedation should be considered in all cases and regarded as mandatory for endoscopic SEMS placement.^17,22 In this procedure, a hydrophilic exchange guide wire and catheter are inserted through the mouth and then negotiated through the stricture. The authors of this present study routinely use a specialized Song-Lim catheter (S&G Biotech, Seongnam, Korea) rather than conventional angiographic catheters for the following reasons: (i) it is easier to manipulate the guide wire through the catheter because there is less friction between the outer surface of the guide wire and inner lumen of the catheter, (ii) it is easier to advance the catheter through the stricture because of the excellent flexibility and pushability of the catheter, and (iii) it allows for the injection of contrast medium through the side arm of the catheter while the guide wire remains in place (Fig. 4).²³ The location and length of the stricture can be identified by injecting a limited amount of water-soluble contrast medium through the catheter, and should be marked securely with radiopaque-markers on the skin of the patient. Next, the hydrophilic exchange guide wire is replaced with a super-stiff guide wire and the catheter is removed. Predilation is generally discouraged to avoid perforation, but can be useful for tight strictures to allow advancement of the delivery system.²⁴ The delivery system is advanced through the stricture over a super-stiff guide wire and the SEMS is deployed under continuous fluoroscopic monitoring. The authors use a 60° J-shaped guiding sheath (S&G Biotech) to facilitate advancement of the delivery system through the stricture when either buckling or loop formation occurs in the remnant stomach, which is often distended (Fig. 5).²⁵ It is important to “overstent” the stricture by ~1.5 to 2 cm on each end to account for stent foreshortening and reduce the risk of tumor overgrowth.⁶ Postdilation can be considered if the stent does not expand to at least half of its fully expanded diameter. Finally, an upper gastrointestinal contrast study is performed the following day to confirm patency of the stent.

Simultaneous self-expandable metal stent placement

The simultaneous placement of two SEMSs is required when a stricture is located at both the afferent and efferent loops in patients with Billroth II reconstruction (Fig. 6).^26,27 Guide wires are placed across both strictures in the afferent and efferent loop. The location and length of the strictures are identified by injection of contrast medium and the strictures marked with radiopaque-markers on the patient’s skin. When it becomes difficult to differentiate the efferent and afferent loops, the guide wire and catheter in each loop are advanced as distally as possible to confirm the loop entered. Alternatively, water soluble contrast medium can be injected through a percutaneous transhepatic biliary drainage (PTBD) tube to opacify the afferent loop.²⁸ After injection, the hydrophilic exchange guide wires are replaced with super-stiff guide wires and the catheters are removed. The delivery systems are advanced through the stricture in the efferent and afferent loops, respectively, over the super-stiff guide wires, allowing the SEMSs to be sequentially deployed under continuous fluoroscopic monitoring. Finally, an upper gastrointestinal series is performed to confirm patency of the stents. Notably, placement of SEMS in the efferent loop alone is sufficient if the patient has a jejunojejunal anastomosis as the afferent loop can drain through the anastomosis (Fig. 7).²⁶ However, the patency of the jejunojejunal anastomosis should be confirmed by negotiating the guidewire and catheter through the anastomosis and injecting water soluble contrast medium.

Percutaneous self-expandable metal stent placement

Generally, it is technically difficult to place a SEMS using the peroral route when a stricture is located at the proximal or middle portion of the afferent loop. In such cases, percutaneous SEMS placement can be performed transhepartically after creation of a PTBD tract under fluoroscopic guidance (Fig. 8).²⁸ Direct percutaneous stent placement into the afferent has been described, but requires the obstructed bowel to lie accessible in the periphery of the abdomen.^29,30 An antibiotic should be administered before SEMS placement and continued for 5 days after the procedure to prevent bile contamination of small bowel bacteria and biliary sepsis, triggered by the procedure. A hydrophilic exchange guide wire is inserted through the PTBD tube and negotiated through the stricture. The PTBD tube is exchanged with a catheter and advanced through the stricture over the guidewire. The stricture location and length are identified by injecting a limited amount of water soluble contrast medium through the catheter, which is then radiopaque-marked on the skin of the patient. Next, the hydrophilic exchange guide wire is replaced with a super-stiff guide wire and the catheter is removed. The PTBD tract is dilated with dilators to allow for advancement of the larger enteral (10–11.5 Fr) stent delivery system, which will require additional analgesia. The SEMS is deployed in the same fashion as for the peroral SEMS placement. A PTBD tube is placed into the common bile duct over the guide wire, and water-soluble contrast medium is injected through the tube to confirm stent patency.

The accepted indications for SEMS placement include malignant gastric-outlet, duodenal, and jejunal obstructions caused by unresectable cancer. The only absolute contraindication for SEMS placement is evidence of gastrointestinal perforation. Mild dysphagia symptoms, distal gastrointestinal obstructions, and terminal-stage disease with a life expectancy < 1 month are relative contraindications. Peritoneal carcinomatosis alone is generally not considered to be a contraindication for SEMS placement;³¹ however, recent studies have shown that patients with carcinomatosis along with ascites are associated with lower clinical success rates,^10,32 most likely because of impaired gastrointestinal motility.^33,34

Studies

A review of the literature revealed a total of 13 studies related to the placement of SEMS for the palliative treatment of recurrent cancer in patients with the surgically-altered stomach.^{26–28,35–44} All studies were retrospective in design, and 6 of the 13 studies were published by our group.^{26,28,37,38,43,44} Additionally, 5 studies included patients with various types of malignancy and/or non-recurrent cancer,^{26–28,41,44} 2 only included patients with esophagojejunostomy,^37,43 1 only included patients with gastroduodenostomy,³⁸ and 2 only included patients with gastrojejunostomy.^26,27 An overview of the literature regarding SEMS placement for the palliative treatment of recurrent cancer in patients with the surgically-altered stomach is presented in Table 1.^{26–28,35–44}

Esophagojejunostomy

In 2007, Kim et al³⁷ reported the outcomes of fluoroscopic SEMS placement for recurrent cancer after total gastrectomy with esophagojejunostomy in 32 patients. In this study, the stricture was located at the anastomotic site in 27 cases (84%) and at the jejunal loop in 5 patients (16%). Technical success was achieved in 30 patients (94%). A retrievable fully-covered nitinol stent (Ni-ti-S Esophageal; Taewoong, Goyang, Korea) (Fig. 9A) was placed in 26 cases (87%) and a partially-covered nitinol stent (Hercules SP Pyloric; S&G Biotech) (Fig. 9B) was placed in 4 patients (13%). In 2 patients (6%), the guidewire or stent delivery system could not pass through the stricture because of complete obstruction and acute angulation of the jejunal loop. Clinical success was achieved in 29 patients (97%). There was 1 patient (3%) who experienced pain immediately after stent placement and required stent removal because of ineffective analgesia. Other complications included pain that was relieved by analgesics in 4 patients (13%), tumor overgrowth in 4 patients (13%), stent migration in 3 patients (10%), and abutment of the tortuous jejunal wall by the end of the stent in 2 patients (6%). The median patency and survival were 87 and 140 days, respectively. These findings indicated that SEMS placement is effective and safe in patients with anastomotic recurrence of gastric cancer after total gastrectomy with esophagojejunostomy. However, no conclusions could be drawn about the outcomes of SEMS placement for jejunal loop obstruction, despite the apparently poor results (technical failure, 40%; complication rate, 33%) because of the small sample size (n = 5).

Subsequently, in 2012, Park et al⁴³ reported the outcomes of fluoroscopic SEMS placement for recurrent cancer at the jejunal loop after total gastrectomy with esophagojejunostomy in 21 patients. Technical success was achieved in 20 patients (95%). A Niti-S Esophageal stent was placed in 10 patients (50%), while the Hercules SP Pyloric stent was placed in 10 patients (50%). In 1 patient (5%), the guidewire could not pass through the stricture because of complete obstruction and acute angulation of the jejunal loop. Clinical success was achieved in 19 patients (95%). There was 1 patient (5%) with two overlapping stents who showed no improvement of symptoms because of impaired motility. Complications occurred in 7 patients (35%), including stent migration (all fully covered stents) in 3 cases (15%), tumor overgrowth in 3 cases (15%), and pain relived by analgesics in 1 patient (5%). The median patency and survival were only 46 and 114 days, respectively. The decreases in the overall patency and survival rates, in comparison with gastroduodenal obstruction or anastomotic strictures, are caused by multiple metastases from disease progression and by peritoneal carcinomatosis with multiple small bowel obstructions in study patients. However, these findings suggest that SEMS placement is effective and safe for the palliation in patients with recurrent cancer at the jejunal loop after total gastrectomy with esophagojejunostomy. Nonetheless, it was obvious that stents with high flexibility and conformability are more appropriate for such cases because of the often tortuous anatomy of the jejunal loop.

Gastroduodenostomy

In 2007, Yang et al³⁸ reported the outcomes of fluoroscopic SEMS placement for recurrent cancer after distal gastrectomy with gastroduodenostomy in 16 patients. In this study, strictures were located at the anastomotic site in all patients. Technical success was achieved in all cases. A Hercules SP Pyloric stent was placed in 11 patients (69%), while a Niti-S Esophageal stent was placed in 5 patients (31%). Clinical success was achieved in 13 patients (81%). There were 2 patients (13%) who could not tolerate any oral intake because of a motility disorder and 1 patient (6%) who refused oral intake because of severe anorexia. Complications included mechanical occlusion of the ampulla of Vater by the covering membrane of the stent, which lead to jaundice in 2 patients (13%), tumor ingrowth resulting from disruption of the polyurethane covering membrane of the stent in 1 patient (6%), stent migration in 1 patient (6%), and stent collapse in 1 patient (6%). Patency was maintained in all but 2 patients (13%) until death (median survival, 52 days). The high technical success rate in this study could be attributed to the relatively obtuse anatomical angulation of the gastroduodenostomy. However, there is a risk of developing jaundice after covered stent placement in such cases because of the close proximity between the gastroduodenal anastomosis and ampulla of Vater.

Gastrojejunostomy

In 2007, Song et al²⁶ reported the outcomes of fluoroscopic SEMS placement in 39 patients with gastrojejunostomy, which included 32 patients (82%) who had recurrent cancer after distal gastrectomy with gastrojejunostomy (Billroth II reconstruction without JJ, 63%; Billroth II reconstruction with JJ, 31%; Roux-en-Y reconstruction, 6%) and 7 patients (18%) who had palliative gastrojejunostomy for a malignant gastroduodenal obstruction. In these cases, the stricture was located at the remnant stomach in 3 patients (8%), anastomotic site in 5 patients (13%), anastomotic site that extended into the afferent loop in 2 patients (5%), anastomotic site that extended into the efferent loop in 17 patients (44%), and anastomotic site that extended into both loops in 12 patients (31%). Technical success was achieved in all patients. A total of 50 stents were placed, including 26 cases (52%) Hercules SP Pyloric stents, 20 cases (40%) Niti-S Esophageal stents, 2 cases (4%) bare nitinol stents (Song Duodenal; Stentech, Seoul, Korea), and 2 cases (4%) fully-covered nitinol stents (Song; Sooho Meditech, Seoul, Korea). Clinical success was achieved in 35 patients (90%). In 2 patients (5%) with stricture at the anastomotic site that extended into the afferent loop or both loops, the operator had mistakenly only stented the afferent loop, resulting in the worsening of symptoms and aspiration pneumonia. There was 1 patient (3%) who had no improvement of symptoms because of distal small bowel obstruction that resulted from peritoneal carcinomatosis. Another patient (3%) who underwent simultaneous SEMS placement showed no improvement of symptoms because the proximal end of the stent in the afferent loop obstructed the passage of food into the stent that was in the efferent loop. Other complications included stent migration in 4 patients (10%), tumor in-/over-growth in 2 patients (5%), mechanical occlusion of the afferent loop by the covering membrane of the efferent loop stent that lead to afferent loop syndrome in 1 patient (3%), and mucosal prolapse in 1 patient (3%). The median patency and survival were 89 and 180 days, respectively. These findings showed the following: (i) SEMS placement is effective and safe in patients with gastrojejunostomy, (ii) accurate knowledge of the surgically-altered anatomy and location of the stricture are critically important for successful treatment outcomes, and (iii) the proximal end of the stent in the efferent loop should be positioned above the stent in the afferent loop, or else the passage of food into the stent in the efferent loop may be obstructed.

Recently, Soo et al²⁷ reported the outcomes of endoscopic simultaneous SEMS placement in the afferent and efferent loops for recurrent cancer in 24 patients with Billroth II reconstruction (43%) or pancreaticoduodenectomy (57%). In this study, bare nitinol stents (Wallflex Duodenal; Boston Scientific, Natick, MA, USA) were used in all patients. The rates of technical and clinical success (96% and 83%, respectively) and complications (30%) were comparable to those reported by Song et al.²⁶ Notably, 2 patients (9%) underwent reintervention because passage of food into the stent in the efferent loop was obstructed by the proximal end of the stent in the afferent loop; one of these patients underwent placement of an additional stent coaxially into the stent in the efferent loop, while the other had the proximal end of the stent in the afferent loop trimmed with argon plasma coagulation. These findings support the recommendations of Song et al²⁶ that the proximal end of the stent in the efferent loop should be positioned above the stent in the afferent loop.

Overall outcomes

In 2001, Park et al³⁵ reported the outcomes of fluoroscopic SEMS placement for recurrent cancer after distal gastrectomy with gastroduodenostomy (18%) or gastrojejunostomy (82%) in 11 patients. In this study, two types of stents were used: a fully-covered stainless-steel Z-stent (Choo Esophageal; Solco Intermed, Seoul, Korea) and a retrievable fully-covered nitinol stent (Song; Doosung Meditec, Seoul, Korea). The technical and clinical success rates were 87% and 82%, respectively. There were no complications, except for 7 patients (64%) with mild to moderate epigastric pain. The high rate of pain reported in this study may be attributed the poor flexibility and conformability of the Choo Esophageal stent, which was used in most of the patients (64%).

In 2004, Jeong et al³⁶ reported the outcomes of fluoroscopic SEMS placement for recurrent cancer in 25 patients with either esophagojejunostomy (46%) or gastrojejunostomy (58%). In this study, three types of stent was used: a fully-covered stainless-steel Z-stent (Choo Esophageal; M.I. Tech, Seoul, Korea), a fully-covered nitinol stent (Niti-S Esophageal; Taewoong), and the retrievable fully-covered Song stent. The technical and clinical success rates were 96% and 100%, respectively. Complications occurred in 7 patients (29%), with stent malfunctions (tumor overgrowth, 17%; stent migration, 4%) accounting for 5 of the cases (71%). Pain only occurred in 2 patients (8%), most likely because the more flexible and conformable nitinol stent was used in most of the cases (63%).

In 2009, Cho et al⁴⁰ reported the outcomes of endoscopic SEMS placement for recurrent cancer after either total gastrectomy with esophagojejunostomy (50%) or distal gastrectomy with gastroduodenostomy (10%) or gastrojejunostomy (40%) in 20 patients. In this study, two types of stent were used: a bare nitinol stent (Niti-S Pyloric; Taewoong) and a fully-covered nitinol stent (Choo Pyloric; M.I. Tech). Technical success was achieved in all patients, although only 14 patients (70%) achieved clinical success. There were 6 patients (30%) who showed no improvement of symptoms because of distal small bowel or colon obstruction (20%), a motility disorder (5%), or incomplete stent expansion (5%). Complications occurred in 7 patients (35%), including stent migration in 3 patients (15%), tumor overgrowth in 2 patients (10%), tumor ingrowth in 1 patient (5%), and incomplete stent expansion in 1 patient (5%). The median patency and survival were 56 and 83 days, respectively. The relatively low clinical success rate observed in that study underlines the importance of careful screening for any distal small bowel or colon obstruction. In the author’s institution, a negative abdominal computed tomography or magnetic resonance imaging is mandatory for SEMS placement.

Recently, in the largest series reported to date, the outcomes of fluoroscopic SEMS placement in 196 patients with the surgically-altered stomach (total gastrectomy with esophagojejunostomy, 37%; distal gastrectomy with gastroduodenostomy, 20%; distal gastrectomy with gastrojejunostomy, 36%; palliative gastrojejunostomy, 7%) were reported by Park et al.⁴⁴ In this study, most (86%) patients had recurrent cancer. The stricture was located at the afferent loop in 14 patients (7%), efferent loop in 135 patients (69%), both loops in 22 patients (11%), and at the jejunal loop in 25 patients (13%). The rates of technical and clinical success were 98% and 96%, respectively. The Hercules SP Pyloric stent was the most frequently used stent (64%); the retrievable fully-covered Niti-S Esophageal stent, retrievable fully-covered Song stent, or fully-covered Song stent were used for the remaining patients (36%). Complications occurred in 48 patients (25%), including stent migration in 21 patients (11%), tumor overgrowth in 15 (8%), pain in 7 patients (4%), stent collapse in 3 patients (2%), and bleeding in 2 patients (1%). The rate of stent migration was lower for the Hercules SP Pyloric stent than for the other types of stents (10% vs 1%; P < 0.001). The median patency and survival were 90 and 131 days, respectively. Adjuvant chemotherapy was associated with increased stent migration (9% vs 2%; P = 0.005) and survival (median, 202 vs 98 days; P < 0.001). These findings indicated that SEMS placement is effective and safe for recurrent cancer in patients with the surgically-altered stomach. Additionally, our findings established that partially-covered stents are less prone to stent migration than fully-covered stents, particularly in cases of patients who had received adjuvant chemotherapy.

Because the survival of patients with recurrent cancer is usually limited to only a few months, an ideal palliative therapy should provide rapid and durable relief of symptoms, result in few complications, require a minimal hospital stay, and prolong patient survival. However, those patients who undergo SEMS placement often fail to achieve the long-term relief of symptoms because of stent malfunction and have to be admitted for reintervention.^42,44 Additionally, palliation with SEMS only provides symptom relief, but does not address the underlying issue of tumor growth. A recent multicenter randomized control trial showed that the placement of a SEMS loaded with radioactive seeds could result in the more durable relief of symptoms and a modest prolongation of survival in patients with unresectable esophageal cancer.⁴⁵ This type of SEMS may also have a role in the palliation of recurrent cancer. Several drug-eluting SEMSs are also under development that has shown the potential to inhibit tumor growth in animals.^46–48

Endoscopic or fluoroscopic SEMS placement is an effective and safe method for the palliation treatment of recurrent cancer in patients with the surgically-altered stomach and should therefore be considered as the first-line option. Collaborations between gastroenterologists and interventional radiologists, as well as a combined endoscopic and fluoroscopic technique, may be necessary for the patients’ benefit as well as further development of the stent technology.

However, accurate knowledge of the surgically-altered anatomy and stricture location are critically important for successful treatment outcomes. In patients with a Billroth II reconstruction in which the stricture is located at both efferent and afferent loops, the proximal end of the stent in the efferent loop should be positioned above the stent in the afferent loop, or else the passage of food into the stent in the efferent loop may be obstructed. Radioactive and drug-eluting SEMSs may potentially represent the future of palliative treatment for surgically-altered stomach.