Types of cancer associated with GOO

GOO is commonly encountered in the context of a variety of cancers including pancreatic, gastric, duodenal, ampullary, cholangiocarcinoma, hepatobiliary, metastatic, and recurrent malignancy following surgical resection. One study involving 176 patients treated at 4 major centers over 7 years cited primary pancreatic malignancy as the most common reason for GOO, corroborating findings from other investigators. Gastric malignancies were the next most common reason for GOO. Other studies have supported pancreatic malignancy as being the most common site to require and benefit from the use of an enteral stent, although some studies have had conflicting data, with gastric malignancy being the most common, followed by pancreatic malignancies. Metastatic cancers to the bowel itself, such as colon cancer, bulky lesions, or adenopathy at the porta hepatis, are also common causes of GOO. Duodenal, ampullary, cholangiocarcinoma, hepatobiliary, and recurrent malignancies are usually less common in incidence than either pancreatic or gastric malignancies, but have been repeatedly reported.

Symptoms and diagnosis of GOO

Symptoms of GOO include nausea, vomiting, weight loss, dehydration, hypoalbuminemia, jaundice, and the inability to tolerate oral intake. Symptoms of GOO can overlap with symptoms of the primary malignancy and may be missed by clinicians, or be attributed to different treatments including chemotherapy and/or radiation.

Diagnosis is based on history, physical examination, results of imaging studies, and findings at endoscopy. A history suggestive of vomiting undigested food hours after eating and nonbilious emesis is strongly suggestive of GOO: these findings imply that food cannot leave the stomach and that the second portion of the duodenum has been isolated from the proximal stomach, respectively.

Physical examination in these patients can be of value, and one may elicit a succession splash on auscultation of the upper abdomen, similar to that seen in children with pyloric stenosis. This maneuver has a sensitivity of 48% if a splash is heard more than 3 hours following ingestion of food.

Imaging techniques vary widely and include plain abdominal radiographs, ultrasonography, upper gastrointestinal (UGI) contrast studies, and abdominal computed tomography (CT). Plain films may show an enlarged gastric bubble, dilated proximal duodenum, and limited air in the small bowel. UGI contrast studies can be performed with either barium or water-soluble contrast. These tests will often demonstrate the location and severity of the site of obstruction, and can be particularly helpful if GOO is incomplete. CT studies may also show a distended stomach, dilated proximal duodenum, decompressed distal small bowel, and the obstructing lesion itself.

The gold standard for the diagnosis of GOO is endoscopy ( Fig. 1 ). Endoscopy is not recommended until appropriate history, physical, and imaging studies are completed. Upper endoscopy can then be performed allowing direct visualization and assessment of severity, nature, site, and biopsy of the obstruction. If the stricture can be traversed with the endoscope this is helpful, but in practice is rare. Fluoroscopy should be used to enhance visibility and safety during the entire procedure. Deployment of an enteral stent involves first traversing the stricture with a soft flexible guide wire. Some endoscopists prefer to then advance a catheter across the stricture and then exchange the soft wire for a stiffer guide wire, but there is no universally accepted preference for type of guide wire. The catheter can also be used to inject contrast across the stricture to fully delineate its length and geometry, which greatly aids in the selection of the proper stent size with regard to both diameter and length. The catheter is then removed, leaving the guide wire in place across the stricture. The stent is then passed through the working channel of the endoscope over the guide wire, and deployed using both direct visualization and fluoroscopic assistance. Most experts recommend the use of fluoroscopy even if the stricture can be traversed without difficulty.

Endoscopic image of gastric outlet obstruction (GOO) at the level of the apex of the duodenal bulb secondary to pancreatic cancer. Note the pinpoint lumen at the center of the image.

Symptoms and diagnosis of GOO

Symptoms of GOO include nausea, vomiting, weight loss, dehydration, hypoalbuminemia, jaundice, and the inability to tolerate oral intake. Symptoms of GOO can overlap with symptoms of the primary malignancy and may be missed by clinicians, or be attributed to different treatments including chemotherapy and/or radiation.

Diagnosis is based on history, physical examination, results of imaging studies, and findings at endoscopy. A history suggestive of vomiting undigested food hours after eating and nonbilious emesis is strongly suggestive of GOO: these findings imply that food cannot leave the stomach and that the second portion of the duodenum has been isolated from the proximal stomach, respectively.

Physical examination in these patients can be of value, and one may elicit a succession splash on auscultation of the upper abdomen, similar to that seen in children with pyloric stenosis. This maneuver has a sensitivity of 48% if a splash is heard more than 3 hours following ingestion of food.

Imaging techniques vary widely and include plain abdominal radiographs, ultrasonography, upper gastrointestinal (UGI) contrast studies, and abdominal computed tomography (CT). Plain films may show an enlarged gastric bubble, dilated proximal duodenum, and limited air in the small bowel. UGI contrast studies can be performed with either barium or water-soluble contrast. These tests will often demonstrate the location and severity of the site of obstruction, and can be particularly helpful if GOO is incomplete. CT studies may also show a distended stomach, dilated proximal duodenum, decompressed distal small bowel, and the obstructing lesion itself.

The gold standard for the diagnosis of GOO is endoscopy ( Fig. 1 ). Endoscopy is not recommended until appropriate history, physical, and imaging studies are completed. Upper endoscopy can then be performed allowing direct visualization and assessment of severity, nature, site, and biopsy of the obstruction. If the stricture can be traversed with the endoscope this is helpful, but in practice is rare. Fluoroscopy should be used to enhance visibility and safety during the entire procedure. Deployment of an enteral stent involves first traversing the stricture with a soft flexible guide wire. Some endoscopists prefer to then advance a catheter across the stricture and then exchange the soft wire for a stiffer guide wire, but there is no universally accepted preference for type of guide wire. The catheter can also be used to inject contrast across the stricture to fully delineate its length and geometry, which greatly aids in the selection of the proper stent size with regard to both diameter and length. The catheter is then removed, leaving the guide wire in place across the stricture. The stent is then passed through the working channel of the endoscope over the guide wire, and deployed using both direct visualization and fluoroscopic assistance. Most experts recommend the use of fluoroscopy even if the stricture can be traversed without difficulty.

Endoscopic image of gastric outlet obstruction (GOO) at the level of the apex of the duodenal bulb secondary to pancreatic cancer. Note the pinpoint lumen at the center of the image.

Nonstent-based treatments for GOO

The classic palliative management of GOO has been either open or laparoscopic gastrojejunostomy (with or without concomitant biliary bypass). Other treatments such as radiation, chemotherapy, nasoenteric feeding tubes, gastric decompression tubes, and total parental nutrition have been used independently or in combination with enteral stent placement and/or gastrojejunostomy.

Laparoscopic surgical gastrojejunostomy is an excellent option, and has been shown to reduce hospital stay and associated morbidity when compared with open gastrojejunostomy. Unfortunately, many patients with GOO are poor surgical candidates because of their advanced underlying disease, poor nutritional status, and short life expectancy. Gastrojejunostomy is performed by attaching a sufficient length of a loop of jejunum to the antecolic portion of the greater curvature of the stomach using a staple gun and/or suture. If the patient has a concomitant biliary stricture, this can be treated via the creation of a surgical biliary bypass or a biliary stent placed endoscopically if the second portion of the duodenum is accessible beyond the site of the GOO. In one study of 23 patients undergoing open gastrojejunostomy, 15 were found to have a biliary stricture, 10 underwent surgical bypass, and 5 had a biliary stent placed endoscopically.

Radiation therapy is also used in the treatment of malignant GOO. Studies investigating monotherapy with radiation have found an increased incidence of tumor growth and fibrosis with worsening obstruction. Mogavero and colleagues found that GOO developed in 7 of 63 patients treated with radiation following curative resection or stent placement for cholangiocarcinoma. However, other studies have demonstrated clinical success with radiation therapy combined with enteral stent placement. In a study by Park and colleagues, radiation therapy following enteral stent placement prolonged stent patency (odds ratio [OR] 0.221, 95% confidence interval [CI] 0.055–0.884, P = .033). By contrast, Im and colleagues found no significant association between stent patency and radiation in 16 patients.

Chemotherapy is also used in the treatment of GOO, usually in combination with another treatment modality. There are conflicting data on whether chemotherapy is helpful in maintaining stent patency. In a study by Cho and colleagues of 27 patients who received chemotherapy, stent patency was prolonged in comparison with those who did not receive chemotherapy (OR 0.34, 95% CI 0.13–0.91, P = .03). Other studies have shown similar results, with chemotherapy prolonging stent patency. By contrast, Im and colleagues found no association between chemotherapy and stent patency in 6 patients who received palliative chemotherapy. Chemotherapy following stent placement has also been associated with increased stent migration in comparison with no other palliative treatment or chemoradiation (42.9% vs 10% vs 9.1%, P = .042).

Nasoenteric tubes are used to treat GOO when life expectancy is very short, typically days to a few weeks. Nasoenteric tubes in patients with short life expectancies are often used as a comfort measure to provide hydration, sometimes in combination with percutaneous endoscopic gastrostomy tubes (for gastric decompression). Nasoenteric tubes should not be left for longer than 1 month. If it is expected that the patient will survive longer than 1 month, enteral stent placement or gastrojejunostomy should be performed. A benefit seen with nasoenteric tubes includes the possibility of removal.

Total parental nutrition (TPN) is often used for the same reasons as nasoenteric tubes. TPN has been used as a temporizing measure before surgery to improve nutritional status through hydration, correction of electrolyte imbalances, improvement in serum albumin levels, and so forth. TPN can be started immediately on recognition of GOO, and continued at home.

Types of enteral stents

In general, enteral stents are composed of differing types of metal alloys with different lengths, diameter, and expansile forces following deployment. Only uncovered stents are currently available in the United States. All currently available enteral stents are designed for through-the-scope deployment and require a therapeutic channel endoscope.

The first stent available for palliation of malignant GOO in the United States was the Enteral Wallstent (Boston Scientific, Natick, MA, USA), which is composed of an uncovered Elgiloy (stainless steel). Its successor, the Enteral Wallflex (Boston Scientific), is available as a nitinol (nickel-titanium) alloy; it has a flared proximal end designed for increased flexibility and decreased migration following deployment. Whereas the Wallstent has bare wire ends, the Wallflex has looped wire ends. There is no evidence that either of these wire ends is superior.

Stents available outside the United States include the Choostent, Hanarostent (MI Tech, Seoul, Korea), and Song duodenal stent (Stentech, Seoul, Korea). The Choostent is made of a flexible expansible metal and is unique in design in that the central portion of the stent is covered in polyurethane, whereas the uncovered flared ends are not. The Hanarostent is available as an uncovered or partially covered self-expanding nitinol stent. The Song stent is similar to the Choostent in that it has a central polyurethane-covered portion with flared ends.

There are many other stents being developed. One example is a design proposed by Kim and colleagues to prevent migration, which includes using a covered “stent within stent design” the additional application of 3 endoscopic clips to the proximal end helps to anchor the stent in place.

Indications for enteral stent placement

Enteral stents are indicated in patients with confirmed malignant GOO, diagnosed through history, physical examination, imaging, and endoscopy. Enteral stents are frequently employed in patients who are poor surgical candidates with shortened life expectancy, advanced or metastatic disease, significant medical comorbidities, and known anesthesia risk.

Surgery, in comparison with enteral stents, is superior in patients with a longer life expectancy because of lower rates of reobstruction. It is difficult to delineate the survival time preferable for surgery, as it is difficult to predict life expectancy. Adler and Baron have proposed an algorithm for patient selection and treatment options, which has been validated by other studies. The algorithm suggests that if a patient has a resectable malignancy then surgical resection should be attempted. If subsequently found to be unresectable at the time of surgery, but localized, and the patient has a good performance status, then a gastrojejunostomy is probably the best option. If the individual has poor performance status, widespread disease, or a severe medical comorbidity, abdominal CT scan and/or UGI series with small bowel follow-through should be obtained. If there is a single site of obstruction and no evidence of peritoneal carcinomatosis, enteral stent placement is the preferred option. If multiple sites of obstruction and/or peritoneal carcinomatosis are identified, a palliative nasoenteric feeding tube or TPN with gastric decompression would be a suitable option.

Contraindications

Contraindications to the placement of an enteral stent are relatively few, and include multiple gastrointestinal tract obstructions (often seen in patients with peritoneal carcinomatosis), life expectancy less than 2 weeks, and underlying gastric dysmotility. Multiple sites of obstruction from metastatic disease and sites that cannot be reached endoscopically (such as the mid to distal jejunum) are contraindications. If life expectancy is less than 1 month, or there are gastric motility problems including linitis plastica, Wai and colleagues have suggested that placement of a nasoenteric tube with gastric decompression is more beneficial.

Enteral stent efficacy and outcomes

Technical success in placing an enteral stent is defined as adequate positioning and deployment of the stent ( Fig. 2 ). Technical success ranges anywhere from 75% to 100% in various studies. One large review article containing data on 1046 patients receiving enteral stents cited technical success in 96%. Common reasons for technical failure include the inability to pass a guide wire across the stricture, inability to advance the stent catheter across the stricture (rare), failure of the delivery system to release the stent, failure of stent deployment, and migration of the stent during the procedure.

Endoscopic Image of a well-placed, fully deployed enteral stent bridging the pylorus in a patient with GOO secondary to pancreatic cancer.

Clinical success with enteral stent placement is defined as relief of symptoms and/or improvement of oral intake. The GOO scoring system (GOOSS) was created to objectively quantify the patient’s level of oral intake both before and following treatment. While other scoring systems have b created, the GOOSS is considered the standard scoring system for assessing clinical success in response to treatment of GOO via any method. The GOOSS uses a score of 0 for no oral intake, a score of 1 for ability to consume liquids only, a score of 2 for soft solids, and a score of 3 for a low-residue or full diet ( Table 1 ).

Clinical success rates in recent studies range from 63% to 97%. Dormann and colleagues analyzed 32 studies with primary data and reported clinical success in 89% of 589 patients following successful stent placement. Patients in this analysis had a mean GOOSS score of 0.4 before stent placement, which increased to a mean of 2.4 following stent placement. The investigators reported that the mean time to overall resolution of symptoms was 4 days, with a range of 1 to 7 days. Reasons for clinical failure occurred in 65 patients, due to disease progression (61%), early stent migration (20%), and procedural-related causes (15%) including the stent being deployed too far proximally or distally, or from inadequate expansion on deployment. In the remaining patients no cause was cited.

Another large review article by Jeurnink and colleagues examined 44 publications with a total of 1000 patients receiving duodenal stents. Clinical success was 89%. Before intervention GOOSS scores were 0 in 62%, 1 in 33%, and 2 in 5%. Following stent placement GOOSS scores were 0 in 6%, 1 in 22%, 2 in 40%, and 3 in 32%. Most patients were able to tolerate an oral diet within 24 hours of stent placement.

Hospital stays following enteral stent placement range from 0 (outpatient) to 18 days. Average hospital time for 324 patients following enteral stent placement was 7 days. In the United States most enteral stents are placed as an outpatient procedure.