- Email: [email protected]
Gastrostomy-to-gastrojejunostomy Tube Conversion: Impact of the Method of Original Gastrostomy Tube Placement
Gastrostomy-to-gastrojejunostomy Tube Conversion: Impact of the Method of Original Gastrostomy Tube Placement Charles Y. Kim, MD, Mayur B. Patel, MD, Michael J. Miller, Jr, MD, Paul V. Suhocki, MD, Anastasia Balius, MD, and Tony P. Smith, MD
PURPOSE: To determine the outcome of gastrostomy tube–to– gastrojejunostomy tube conversion on the basis of the method of original gastrostomy tube placement. MATERIALS AND METHODS: One hundred twenty-four patients (age range, 13– 87 years; 72 male and 52 female patients) underwent conversion of a primarily placed gastrostomy tube to a gastrojejunostomy tube at the authors’ institution between January 2000 and December 2008. The method of original gastrostomy tube placement was radiologic (n ⴝ 27), endoscopic (n ⴝ 75), laparoscopic (n ⴝ 2), or open surgery (n ⴝ 20). The method of placement was correlated with the success rates of gastrostomy-to-gastrojejunostomy tube conversion. Medical records and radiologic images were reviewed to determine the frequency of proximal migration of the jejunostomy tube into the stomach. Follow-up data were available for an average of 136 days after gastrostomy-to-gastrojejunostomy tube conversion (median, 63 days; range, 1–1,300 days). RESULTS: Of 124 gastrostomy tube–to– gastrojejunostomy tube conversions, 109 (87.9%) were successfully performed. Procedural conversion failure occurred in one of the 27 radiologically inserted gastrostomy tubes (3.7%) compared to 14 of the 97 (14%) nonradiologically inserted gastrostomy tubes (P ⴝ .19), of which 12 were inserted endoscopically and two were inserted surgically. Of the 109 patients with successful tube conversion, jejunal tip malposition occurred at follow-up in 18 (16.5%). Of these, four patients developed aspiration pneumonia (22%), which contributed to patient death in two. The frequency of jejunal tip malposition was 3.8% (one of 26 patients) for radiologically placed gastrostomy tubes and 20% (17 of 83 patients) for nonradiologically placed gastrostomy tubes (P ⴝ .07). Combined, 32% of gastrostomy tubes placed nonradiologically resulted in either procedural failure or eventual jejunal tip malposition, compared to 7.4% of radiologically placed gastrostomy tubes (P ⴝ .01). CONCLUSIONS: The frequency of procedural failure or eventual jejunal tip malposition with conversion of radiologically placed gastrostomy tubes to gastrojejunostomy tubes is significantly lower with radiologically placed gastrostomy tubes than with nonradiologically inserted gastrostomy tubes. J Vasc Interv Radiol 2010; 21:1031–1037
PERCUTANEOUS gastrostomy tubes are often necessary for patients with imFrom the Department of Radiology, Division of Vascular and Interventional Radiology, Duke University Medical Center, Box 3808, Durham, NC 27710 (C.Y.K., M.B.P., M.J.M., P.V.S., T.P.S.); and University Radiology, Knoxville, Tennessee (A.B.). Received April 2, 2009; final revision received October 28, 2009; accepted April 3, 2010. Address correspondence to C.Y.K.; E-mail: [email protected] None of the authors have identified a conflict of interest. © SIR, 2010 DOI: 10.1016/j.jvir.2010.04.003
paired neurologic or gastrointestinal status for the purposes of enteral alimentation and gastric decompression (1–3). However, aspiration pneumonia in the setting of esophageal reflux poses a considerable risk in patients with gastrostomy tubes. The benefit of gastrojejunostomy tube placement is a markedly decreased risk of enteral alimentation reflux by virtue of the jejunal tip being positioned beyond the ligament of Treitz (4). Radiologic conversion of a gastrostomy tube to a gastrojejunostomy tube involves removal of the gastrostomy
tube and insertion of a gastrojejunostomy tube through the preexisting tract, with subsequent negotiation of the jejunal portion of the tube through the pylorus into the duodenum and proximal jejunum by using fluoroscopic guidance. However, it has been suggested that an unfavorable angulation of the gastrostomy tube tract relative to the pylorus may result in failure to cannulate the duodenum (5,6). Furthermore, in the presence of such suboptimal tract angulation, the jejunal portion of the gastrojejunostomy tube may be prone to retrograde migration into the stomach.
1031
1032
•
Gastrostomy-to-gastrojejunostomy Tube Conversion
Accordingly, it has been our experience that some gastrojejunostomy tubes are susceptible to multiple episodes of retraction of the tip of the catheter back into the stomach, requiring multiple repeat procedures to reposition the tip back into the jejunum. Percutaneous gastrostomy tubes can be placed by using open or laparoscopic surgical, endoscopic, or radiologic techniques (7–11). During radiologic placement of gastrostomy tubes, a favorable angulation toward the pylorus is typically used. However, this angulation toward the pylorus can be very difficult to achieve with nonradiologic gastrostomy tube placement. The purpose of this retrospective study was to evaluate whether the method of gastrostomy tube placement affects the success of gastrostomy-to-gastrojejunostomy tube conversion and subsequent jejunal tip malposition with currently used large-bore gastrojejunostomy tubes.
MATERIALS AND METHODS This retrospective study was approved by the institutional review board, and a waiver of the informed consent requirement was obtained for this Health Insurance Portability and Accountability Act– compliant study. Patient Selection and Data Collection From January 2000 through December 2008, 130 patients underwent gastrostomy-to-gastrojejunostomy tube conversion at our institution. Six patients had a previously performed surgical gastrojejunostomy anastomosis and are not included here; therefore, the study cohort consists of 124 adult and pediatric patients. The mean patient age was 60 years (range, 13– 87 years; median, 61 years). There were 72 male and 52 female patients. The gastrostomy tubes were initially inserted radiologically in 27 patients, endoscopically in 75, with open surgery in 20, and laparoscopically in two. The most common primary indication for gastrostomy-to-gastrojejunostomy tube conversion was aspiration or risk of aspiration (n ⫽ 83), followed by gastroparesis (n ⫽ 24), pancreatitis (n ⫽ 6), leakage from the gastrostomy tube exit site (n ⫽ 6), gastric outlet obstruction (n ⫽ 3), gastric perforation (n ⫽ 1), and superior mesenteric artery syndrome (n ⫽ 1). Indications for
original gastrostomy tube insertion included respiratory failure with prolonged intubation (n ⫽ 44), failure to thrive (n ⫽ 28), neurologic deficit with impaired swallowing (n ⫽ 23), head and neck tumor (n ⫽ 9), gastric outlet obstruction (n ⫽ 9), esophageal perforation (n ⫽ 4), aspiration (n ⫽ 4), and severe dysphasia (n ⫽ 3). The clinical, radiologic, and surgical notes were reviewed for procedural success and gastrojejunostomy tube tip position at follow-up. Complications were assessed in accordance with the SIR Reporting Standards (12). Gastrostomy Tube–to– gastrojejunostomy Tube Conversion Technique All procedures were performed in an interventional radiology suite using fluoroscopy. Procedures were performed by an attending interventional radiologist or an interventional radiology fellow under the supervision of an attending radiologist. Conscious sedation with midolazam and fentanyl was used for all patients. The preexisting gastrostomy tube was removed over a guide wire, which was then manipulated into the pylorus and duodenum by using a 5-F angled catheter. If this was unsuccessful, multiple catheters and devices were used to cannulate the pylorus. These included modified catheters, vessel dilators, transjugular liver biopsy sets (Liver Access and Biopsy Set; Cook, Bloomington, Indiana), and transjugular intrahepatic portosystemic shunt sets (Rosch-Uchida Transjugular Liver Access Set, Cook) to increase the chances of successful small bowel catheterization. Once the guide wire was positioned in the proximal jejunum, the catheter was removed and a new 18-F MIC gastrojejunostomy tube (Medical Innovation, Draper, Utah) was inserted over the guide wire with the tip in the fourth portion of the duodenum or beyond the ligament of Treitz. Injection of a small amount of iodinated contrast medium was used to confirm appropriate jejunal tip position. The gastric port was then injected with iodinated contrast medium or air to confirm appropriate intragastric positioning of the gastric vent. The intragastric retention balloon was then inflated with saline, sometimes with a tiny amount of contrast medium. If the guide wire or gastrojejunostomy tube could not be ma-
July 2010
JVIR
nipulated through the pylorus into the duodenum despite multiple attempts, the procedure was aborted and a gastrostomy tube reinserted (Medical Innovation). The former was deemed a successful procedure, with the latter considered a failure of gastrostomy-togastrojejunostomy tube conversion. Radiologic Gastrostomy Tube Insertion Technique Briefly, patients were administered barium orally or via a nasogastric tube at least 12 hours prior to the planned procedure, in order to opacify the transverse colon. After gastric insufflation, a gastrostomy tube site was chosen below the costal margin, above the transverse colon, and to the left of midline. Two percutaneous T-fasteners were guided fluoroscopically into the stomach and deployed on either side of the planned gastrostomy tube incision. A needle was inserted into the stomach angled toward the pylorus, through which a guide wire was inserted. After serial dilation of the tract, an 18-F MIC gastrostomy tube was inserted. Endoscopic Gastrostomy Tube Insertion Technique After advancement of a gastroscope via the mouth into the stomach, the stomach was insufflated with air. Pressure was applied externally in the left upper quadrant two finger breadths below the left costal margin. As the light source was advanced to the area of maximal indentation internally, transillumination through the skin was seen externally. An 18-gauge needle was then passed though the anterior abdominal wall into the stomach under direct vision. A guide wire was passed into the stomach and then grasped with an endoscopic snare and pulled retrograde out the mouth. A gastrostomy tube was then advanced anterograde through the mouth, into the stomach, and out the anterior abdominal wall. Open Surgical Gastrostomy Tube Insertion Technique Open gastrostomy tube insertions were performed in a Stamm fashion. Briefly, a short upper midline incision was made and the peritoneal cavity entered. At a relatively avascular area in the mid-body of the anterior stomach,
Volume 21
Number 7
two silk purse-string stitches are placed. A 24-F gastrostomy tube was inserted through a separate stab incision in the left upper quadrant and into an incision located at the center of the pre-placed purse-string stitches, followed by balloon inflation and tightening of the purse-string stitches. The stomach wall adjacent to the tube insertion site was then sutured to the anterior abdominal wall to allow solid fusion to take place and to prevent leakage into the peritoneal cavity. Laparoscopic Gastrostomy Tube Insertion Technique A laparoscope was inserted via a 1-cm midline incision and the peritoneal cavity insufflated with air. A 1-cm incision was made in the left upper quadrant for placement of a 5-mm port. The stomach was insufflated with air via a nasogastric tube and then retracted toward the left upper quadrant port site. The stomach was then fixed to the anterior abdominal wall by using four Tfasteners or nylon sutures between the stomach and anterior abdominal wall under laparoscopic visualization. Once the stomach was secured to the abdominal wall, a needle was advanced into the stomach, through which a guide wire was inserted. In a Seldinger fashion, the wire was exchanged with a series of enlarging dilators. Once the tract was sufficiently dilated, a gastrostomy tube was inserted through a peel-away introducer sheath. Statistical Methods Data analysis was performed using software (SAS version 9; SAS Institute, Cary, North Carolina). Statistical analysis of technical failure and malposition rates was performed and correlated with the method of original gastrostomy placement, indication for initial gastrostomy tube insertion, indication for gastrostomy-to-gastrojejunostomy tube conversion, and length of time between gastrostomy insertion and conversion by using the Fisher exact test. A P value of less than or equal to .05 was considered statistically significant.
RESULTS Of the 124 patients referred for gastrostomy-to-gastrojejunostomy tube conversion with a surgically nonaltered
Kim et al
•
1033
Table 1 Frequency of Procedural Failures and Jejunal Catheter Tip Malposition according to Method of Gastrostomy Placement in Stomachs without Surgical Alteration Gastrostomy Method Radiologic Endoscopic Open surgical Laparoscopic Nonradiologic Total
No. of Patients
Procedural Failure
Jejunal Tip Malposition*
Failure or Malposition
27 75 20 2 97 124
1 (3.7) 12 (16) 2 (10) 0 (0) 14 (14) 15 (12.1)
1 (3.8) 14 (22) 3 (17) 0 (0) 17(20) 18 (16.5)
2 (7.4) 26 (35) 5 (25) 0 (0) 31 (32) 33 (26.6)
Note.—Numbers in parentheses are percentages. * Number of patients for percentage calculation equals number of patients minus patients with procedural failure.
stomach, 15 (12.1%) had procedural failures due to inability to manipulate a guide wire into the duodenum (Table 1). One of the 27 radiologically placed gastrostomy tubes (3.7%) resulted in failed conversion to a gastrojejunostomy tube due to inability to cannulate the pylorus, whereas 14 of the 97 nonradiologically placed gastrostomy tubes (14%) resulted in conversion failure (P ⫽ .19); (Figs 1, 2). Follow-up data were available for an average of 136 days after gastrostomy tube–to– gastrojejunostomy tube conversion (median, 63 days; range, 1–1,300 days). Endpoints for follow-up included the following: lost to follow-up (n ⫽ 39), death (n ⫽ 30), gastrojejunostomy tube removal because no longer needed (n ⫽ 23), currently in place (n ⫽ 5), infection (n ⫽ 3), conversion to a gastrostomy tube due to recurrent malposition (n ⫽ 3), gastrojejunostomy tube fell out (n ⫽ 2), de novo gastrojejunostomy tube placement at new site due to recurrent malposition (n ⫽ 1), and gastrojejunostomy tube removal due to exit-site excoriation (n ⫽ 1). Of the 109 patients with successful gastrostomy tube–to–gastrojejunostomy tube conversion, 18 had eventual retrograde malposition of the jejunal catheter into the stomach. The average interval between gastrostomy tube conversion and jejunal catheter malposition was 116 days (median, 47 days; range, 5–1,059 days). The average number of subsequent jejunal catheter repositioning procedures in each of these 18 patients was 1.8 (median, 1; range, 1–10). Correlation with the method of initial gastrostomy tube placement revealed that one of the 26 (3.8%) radiologically placed gastrostomy tubes that were successfully con-
verted to gastrojejunostomy tubes had eventual jejunal tip malposition compared to 17 of the 83 (20%) nonradiologically placed gastrostomy tubes that were successfully converted to gastrojejunostomy tubes (P ⫽ .07). The final disposition of the 18 malpositioned gastrojejunostomy tubes was as follows: three were removed because they were no longer needed, two were converted to a gastrostomy tube, one was replaced with a new gastrojejunostomy tube inserted via a de novo puncture, and one was removed and replaced with a laparoscopically inserted jejunostomy tube. In the remaining 11 patients, the patient either died (n ⫽ 6) or was lost to follow-up (n ⫽ 4), or the gastrojejunostomy tube is currently in use (n ⫽ 1). Four of the 18 patients with malpositioned gastrojejunostomy tubes had aspiration pneumonia that was directly attributed to continued tube feeding through a gastrojejunostomy tube, which had migrated proximally into the stomach or esophagus (Table 2). In three of these cases, the tip of the gastrojejunostomy tube was found to be in the esophagus (Fig 3). In two of the patients with gastrojejunostomy tube tip malposition, the clinical team had deemed that the tube malposition-related aspiration pneumonia contributed to the death of the patient. One patient was admitted four times due to aspiration pneumonia associated with recurrent gastrojejunostomy tube tip malposition. Overall, 7.4% of radiologically inserted gastrostomy tubes resulted in gastrostomy tube conversion failure or eventual malposition, compared to 32% of nonradiologically placed gastrostomy tubes (P ⫽ .01). No statistically signifi-
1034
•
July 2010
Gastrostomy-to-gastrojejunostomy Tube Conversion
Figure 1. Favorable tract angulation with radiologic gastrostomy tube insertion. (a) Radiograph obtained after injection of iodinated contrast medium through the radiologically inserted gastrostomy tube (arrows) in a 64-year-old woman who required enteral alimentation due to anoxic brain injury. Note the gastric insertion site at the proximal gastric antrum with orientation of the tract toward the pylorus (*). (b) Axial computed tomographic (CT) scan in a 76-year-old woman after radiologic gastrostomy tube insertion due to poor functional status and malnutrition. Notice the transabdominal tract angulation (arrows) at approximately 40° in relation to the skin toward the pylorus (*).
Figure 2. Unfavorable tract angulation of a gastrostomy tube originally inserted endoscopically. (a) Radiograph obtained after injection of a small amount of iodinated contrast medium through the gastrostomy tube in a 24-year-old man in need of enteral nutrition due to traumatic brain injury and poor functional status. The gastrostomy tube (arrows) had been previously exchanged for a balloon-tip gastrostomy tube radiologically. Note the gastric insertion site at the mid-gastric body, with tract orientation toward the gastric fundus (*). (b) CT scan in an 81-year-old man in need of enteral nutrition due to traumatic brain injury and poor functional status. The gastrostomy tube (arrows) had also been previously exchanged for a balloon-tip gastrostomy tube radiologically. The CT scan demonstrates tract angulation toward the patient’s left side, away from the pylorus. Due to the tract angulation and direction, this gastrostomy tube will likely be difficult to convert to a gastrojejunostomy tube and will likely be prone to retrograde tip malposition into the stomach.
cant differences were found between the nonradiologically inserted gastrostomy tubes. In addition, patients who
were referred for gastrostomy-to-gastrojejunostomy tube conversion with the indication of gastric outlet obstruc-
JVIR
tion or gastroparesis were significantly more likely to experience a technical failure of conversion (25% vs 8%, P ⫽ .05), although there was no significant difference in eventual malposition rates. The indication for initial gastrostomy tube insertion and length of dwell time of the gastrostomy tube before conversion were not significantly correlated with technical success or eventual tip malposition. After gastrostomy tube conversion, exit site infection occurred in three patients. Subsequent catheter removal was performed in all three patients to help resolve the infection. However, in two of these patients, pericatheter leakage of gastric contents and signs and symptoms of exit site infection were present before the gastrostomy tube conversion procedure and, in fact, were the indications for gastrostomy tube conversion. In one of these patients, the catheter was removed on postprocedure day 1, with subsequent de novo gastrostomy tube insertion via laparoscopic surgery. The other patient’s catheter was removed on postprocedure day 8, and a nasogastric tube and nasojejunal feeding tube were inserted to address the patient’s enteral requirements until the patient’s death 3 days later. The third patient’s gastrojejunostomy tube was removed 13 days after gastrostomy tube conversion due to new cellulitis around the gastrojejunostomy tube exit site that developed after the conversion procedure. However, the patient’s gastroparesis had resolved in the interim, and therefore no additional intervention was performed after catheter removal. The only procedural complication was this single exit site infection. No other complications related to gastrostomy tube conversion were identified.
DISCUSSION Both gastric and jejunal tube feeding have proven to be effective methods for providing nutrition to patients who cannot tolerate oral alimentation (13). Whereas jejunal feeding is typically infused in a continuous manner, gastric feeding may have a physiologic benefit and has the option of bolus infusion, which is particularly convenient for the non-bed-ridden patient. At our institution, gastrostomy tube insertion is frequently the initial intervention of choice for these reasons and also because gastrostomy tube insertion is technically
Volume 21
Number 7
Kim et al
Table 2 Complications Associated with Gastrojejunostomy Tube Tip Malposition Event
No. of Patients (n ⫽ 18)
No. of Episodes (n ⫽ 33)
Gastrojejunostomy tube tip in esophagus Aspiration pneumonia Death related to aspiration pneumonia
4 (22) 4 (22) 2 (11)
11 (33) 7 (21) NA
Note.—Numbers in parentheses are percentages. NA ⫽ not applicable.
easier, faster, and less expensive than gastrojejunostomy tube insertion and has fewer episodes of clogging (7,11). However, in the setting of aspiration risk, gastric dysmotility, or pancreatitis, primary gastrojejunostomy tube insertion is indicated. If these conditions occur or are discovered after gastrostomy tube insertion, then gastrostomy-to-gastrojejunostomy tube conversion is indicated. At our institution, gastrostomy tubes are initially inserted endoscopically or radiologically on the basis of referring physician preference or availability. Open gastrostomy tube insertion typically occurs only when a patient undergoes major gastrointestinal surgery, where nontransient enteral feeding requirements are anticipated. During endoscopic gastrostomy tube placement (14), the orientation of the gastrostomy tube tract can be difficult to dictate. The tract, established by means of initial needle insertion, tends to be perpendicular to the skin and stomach due to safety concerns for injury to nonvisible intraabdominal structures, such as the colon, and is sometimes angled toward the gastroesophageal junction. Depending on the anatomy of the stomach, this can create an acute angle for a guide wire or gastrojejunostomy tube to reach the duodenum. Similarly, during open surgical or laparoscopic gastrostomy tube placement (15,16), the tract is also often created relatively perpendicular or randomly in terms of angulation with the skin and stomach. In fact, an angulated tract is often avoided during open surgical gastrostomy tube insertion to allow an adequate fascial edge for closure. Because gastrostomy tube–to– gastrojejunostomy tube conversion is not typically performed by most surgeons or endoscopists, the issue of tract angulation is likely not taken into account during gastrostomy tube placement. Although it is possible to attempt to create
the tract in a favorable orientation toward the pylorus during surgical gastrostomy placement, the final geometric orientation of the tract is difficult to predict because the tract is created under nonanatomic conditions. During open surgery, the gastrostomy incision is made while the stomach is exteriorized through the primary incision and then a separate incision made through the skin for the planned exit site. Therefore, the tract angulation may change considerably once the stomach returns to its natural, in vivo configuration. Similarly, during laparoscopic gastrostomy tube placement, the gastrostomy incision and skin incision are made during peritoneal insufflation, which alters the normal anatomic relationship of the stomach with regard to the planned exit site. Conversely, during radiologic placement of gastrostomy tubes creation of a favorable tract angulation is straightforward to perform. Because the tract is created by needle insertion through the skin into the stomach during real-time fluoroscopic visualization of the stomach in its normal in vivo configuration, an exaggerated angle can be created toward the pylorus, in effect, as much as desired. Furthermore, because the gastric puncture is made with the anterior stomach wall already fixed to the anterior abdominal wall, this angulation will change minimally after completion of the procedure. Finally, because gastrostomy tube conversion is most commonly performed by interventional radiologists at our institution, an appropriate tract orientation is routinely used for gastrostomy tube placement in case later gastrojejunostomy tube conversion is required. On the basis of our experience and that of others (5,6), we believe that suboptimal tract orientation created at the time of initial gastrostomy tube insertion is the primary cause of procedural conversion failures. An unfavorable
•
1035
tract orientation in relation to the gastric lumen and pylorus can make it very difficult or impossible to manipulate a guide wire and catheter into the duodenum due to the poor mechanical vector and tendency for buckling into the fundus of the stomach (Figs 1, 2). In addition, the site of original puncture of the stomach for gastrostomy and anatomic configuration of the stomach and pylorus may also play a role. Although there was a tendency for greater success with conversion of radiologically placed gastrostomy tubes (96.7%) as compared to nonradiologically placed gastrostomy tubes (85.6%), this did not reach statistical significance. However, not surprisingly, gastrostomy conversion procedures with gastroparesis or gastric outlet obstruction as the indication were significantly more likely to result in technical failure. No other definite predictors of success or failure were identified. An unfavorable tract angle is also presumed to be the reason for subsequent retrograde malposition of the jejunal portion of the gastrojejunostomy tube into the stomach. In our series, 16.5% of patients experienced retrograde malposition of the jejunal tip into the stomach and, of these, 22% resulted in the tip in the esophagus, which creates an extremely high aspiration risk even in patients without gastroesophageal reflux (Fig 3). In two of the patients, the associated aspiration pneumonia contributed to their deaths. Therefore, it is important for the operator to identify those patients at risk for gastrojejunostomy tube tip malposition and consider alternate procedures. It is also important for the patients’ physicians to be aware of this substantial risk for malposition with resultant aspiration, perhaps warranting more frequent monitoring with abdominal radiographs to assess gastrojejunostomy tube position in the setting of feeding intolerance. Moreover, repositioning of malpositioned gastrojejunostomy tubes is costly and incurs additional radiation dose to the operator and patient. In these patients with recurrent episodes of tip migration, primary jejunostomy tube insertion may be indicated (17). Alternatively, removal of the gastrostomy tube and creation of a new puncture tract with a more favorable angulation toward the pylorus could be performed for de novo gastrojejunostomy tube insertion. It should be kept in mind that additional possible contribu-
1036
•
Gastrostomy-to-gastrojejunostomy Tube Conversion
Figure 3. Images in a 68-year-old man with locked-in syndrome who required enteral nutrition due to poor functional status. Gastrostomy tube–to– gastrojejunostomy tube conversion was performed due to episodes of aspiration. (a) Fluoroscopic spot image obtained after successful gastrostomy tube conversion after injection of iodinated contrast medium through the jejunostomy and gastric port. Opacification of the proximal jejunum (arrowheads) helps confirm the intrajejunal location of the tip (arrow). (b) Two months later, chest radiograph demonstrates the gastrojejunostomy tube to be coiled in the gastric fundus, with the jejunal tip extending into the proximal esophagus (arrows). This patient had aspiration pneumonia secondary to enteral feeding through this tube. On some of this patient’s chest radiographs, the malpositioned gastrojejunostomy tube was mistaken for a nasogastric tube.
tors to jejunal catheter malposition may include excessive peristalsis, reverse peristalsis, vomiting, gastric anatomy, and anatomic location of gastric puncture. On the basis of our review of the literature, only one article has compared success rates of gastrostomy tube–to– gastrojejunostomy tube conversion on the basis of the method of original gastrostomy tube placement (5). In that study, 38 patients had surgical gastrostomy tube placement (with 18% procedural failure) and 18 had endoscopic gastrostomy tube placement (with 22% procedural failure); only seven had radiologic gastrostomy tube placement (with no procedural failures). No statistical analysis was specified, likely because the number of patients with radiologic gastrostomy tube placement was inadequate to achieve statistical significance. In terms of evaluation of eventual jejunal tip malposition (n ⫽ 8), the mean length of follow-up was only 9 weeks (range, 1–16 weeks). Furthermore, their analysis was based on small-bore (12–16-F) catheters, some or all of which were single-lumen transgastric jejunal feeding catheters, whereas the ones analyzed in our study are the more commonly used large-bore 18-F
gastrojejunostomy tubes with both a gastric and jejunal port. These largerbore catheters tend to be stiffer, which has an unknown effect on the likelihood for buckling into the stomach and subsequent jejunal tip malposition. The primary limitation of this study is its retrospective nature, which prohibited analysis of a number of parameters that may help predict gastrostomy tube conversion failure and eventual gastrojejunostomy tube tip malposition. As mentioned, the authors and others believe that the gastrostomy tube tract orientation has a substantial effect on the outcome of gastrostomy tube conversion (5,6). However, a formal analysis of tract orientation was not possible due to a lack of assessment at conversion. Unfortunately, a very small minority of the patients underwent CT after gastrostomy tube insertion and before conversion to a gastrojejunostomy tube, which would be helpful for objectively assessing tract orientation and measuring tract angles. In addition, the gastric anatomy and actual location of gastric puncture could not be retrospectively assessed due to the variability in abdominal imaging acquisition and variable presence of adequate intraluminal air or contrast
July 2010
JVIR
medium to delineate gastric anatomy. Furthermore, a subjective assessment by the interventional radiologist performing the procedure was inconsistently documented, which may have also provided potentially predictive information. In our series, the interventional radiologists went to great lengths to successfully convert gastrostomy tubes to gastrojejunostomy tubes, using a wide array of interventional tools and prolonged fluoroscopy times, for example, as high as 42 minutes. Unfortunately, procedural details of these additional measures and devices used for gastrostomy tube conversion were inconsistently reported and fluoroscopy times were not documented for most procedures. This information may have allowed for a cost analysis that could potentially affect decision making. Considering the median postgastrostomy tube conversion follow-up of 63 days and median time to jejunal tip malposition of 47 days, 48 patients (44%) had follow-up data for fewer than 47 days. Therefore, the malposition rates in our study are likely underestimated. Although this study represents the largest number of gastrostomy tube–to– gastrojejunostomy tube conversions in the literature, the number of procedures is somewhat limited. With a larger number of patients, both the conversion success rate and jejunal tip malposition rate may have achieved statistical significance. In summary, the results of this study demonstrated that gastrostomy tubes inserted radiologically have improved outcomes when converted to gastrojejunostomy tubes, in terms of procedural failure rates and eventual malposition rates, when compared to nonradiologically inserted gastrostomy tubes. Our results suggest that patients with a high probability for gastrostomy tube conversion may benefit from radiologic placement of the original gastrostomy tube with intentional angulation toward the pylorus. When faced with a patient with a nonradiologically placed gastrostomy tube in need of a gastrojejunostomy tube, a careful assessment of the gastrostomy tube tract orientation in relation to the pylorus is likely warranted. In cases of poor tract angulation or difficulty cannulating the pylorus, primary jejunostomy tube insertion or a de novo gastrojejunostomy tube access with a more favorable orientation should be considered in light of the 26.6% fre-
Volume 21
Number 7
quency of procedural failure or eventual jejunal tip malposition for all procedures. Although overall procedural success rates are high with all methods of gastrostomy tube placement, a dangerous scenario is subsequent occult retrograde catheter tip malposition, which resulted in a 22% frequency of aspiration pneumonia in this largely debilitated patient population. Therefore, operator and referring clinician awareness of this not-infrequent complication should be emphasized. In addition, gastrostomy tube insertion with nonradiologic techniques may benefit from procedural modification to minimize procedural failure and jejunal tip malposition when subsequent conversion to a gastrojejunostomy tube is required. References 1. O’Keeffe F, Carrasco CH, Charnsangavej C, Richli WR, Wallace S, Freedman RS. Percutaneous drainage and feeding gastrostomies in 100 patients. Radiology 1989; 172:341–343. 2. Halkier BK, Ho CS, Yee AC. Percutaneous feeding gastrostomy with the Seldinger technique: review of 252 patients. Radiology 1989; 171:359 –362. 3. van Sonnenberg E, Wittich GR, Cabrera OA, et al. Percutaneous gastrostomy and gastroenterostomy. II. Clinical experience. AJR Am J Roentgenol 1986; 146:581–586.
Kim et al
4. Gustke RF, Varma RR, Soergel KH. Gastric reflux during perfusion of the proximal small bowel. Gastroenterology 1970; 59:890 – 895. 5. Lu DS, Mueller PR, Lee MJ, Dawson SL, Hahn PF, Brountzos E. Gastrostomy conversion to transgastric jejunostomy: technical problems, causes of failure, and proposed solutions in 63 patients. Radiology 1993; 187:679 – 683. 6. McLean GK, Rombeau JL, Caldwell MD, Ring EJ, Freiman DB. Transgastrostomy jejunal intubation for enteric alimentation. AJR Am J Roentgenol 1982; 139:1129 –1133. 7. Barkmeier JM, Trerotola SO, Wiebke EA, et al. Percutaneous radiologic, surgical endoscopic, and percutaneous endoscopic gastrostomy/gastrojejunostomy: comparative study and cost analysis. Cardiovasc Intervent Radiol 1998; 21:324 –328. 8. Silas AM, Pearce LF, Lestina LS, Grove MR, et al. Percutaneous radiologic gastrostomy versus percutaneous endoscopic gastrostomy: a comparison of indications, complications and outcomes in 370 patients. Eur J Radiol 2005; 56:84 –90. 9. Consentini EP, Sautner T, Gnant M, Winkelbauer F, Teleky B, Jakesz R. Outcomes of surgical, percutaneous endoscopic, and percutaneous radiologic gastrostomies. Arch Surg 1998; 133: 1076 –1083. 10. Wollman B, D’Agostino HB, WalusWigle JR, Easter DW, Beale A. Ra-
11.
12.
13. 14.
15.
16.
17.
•
1037
diologic, endoscopic, and surgical gastrostomy: an institutional evaluation and meta-analysis of the literature. Radiology 1995; 197:699 –704. Hoffer EK, Cosgrove JM, Levin DQ, Herskowitz MM, Sclafani SJ. Radiologic gastrojejunostomy and percutaneous endoscopic gastrostomy: a prospective, randomized comparison. J Vasc Interv Radiol 1999; 10:413– 420. Sacks D, McClenny TE, Cardella JF, Lewis CA. Society of Interventional Radiology clinical practice guidelines. J Vasc Interv Radiol 2003; 14: S199 –S202. Jabbar A, McClave SA. Pre-pyloric versus post-pyloric feeding. Clin Nutr 2005; 24:719 –726. Safadi BY, Marks Jm, Ponsky JL. Percutaneous endoscopic gastrostomy. Gastrointest Endosc Clin N Am 1998; 8:551–568. Apelgren KN, Zambos J. Is percutaneous better than open gastrostomy? A clinical study in one surgical department. Am Surg 1989; 55:596 – 600. Murayama KM, Johnson TJ, Thompson JS. Laparoscopic gastrostomy and jejunostomy are safe and effective for obtaining enteral access. Am J Surg 1996; 172:591–594. Tapia J, Murguia R, Garcia G, de los Monteros PE, Onate E. Jejunostomy: techniques, indications, and complications. World J Surg 1999; 23:596 – 602.
PURPOSE: To determine the outcome of gastrostomy tube–to– gastrojejunostomy tube conversion on the basis of the method of original gastrostomy tube placement. MATERIALS AND METHODS: One hundred twenty-four patients (age range, 13– 87 years; 72 male and 52 female patients) underwent conversion of a primarily placed gastrostomy tube to a gastrojejunostomy tube at the authors’ institution between January 2000 and December 2008. The method of original gastrostomy tube placement was radiologic (n ⴝ 27), endoscopic (n ⴝ 75), laparoscopic (n ⴝ 2), or open surgery (n ⴝ 20). The method of placement was correlated with the success rates of gastrostomy-to-gastrojejunostomy tube conversion. Medical records and radiologic images were reviewed to determine the frequency of proximal migration of the jejunostomy tube into the stomach. Follow-up data were available for an average of 136 days after gastrostomy-to-gastrojejunostomy tube conversion (median, 63 days; range, 1–1,300 days). RESULTS: Of 124 gastrostomy tube–to– gastrojejunostomy tube conversions, 109 (87.9%) were successfully performed. Procedural conversion failure occurred in one of the 27 radiologically inserted gastrostomy tubes (3.7%) compared to 14 of the 97 (14%) nonradiologically inserted gastrostomy tubes (P ⴝ .19), of which 12 were inserted endoscopically and two were inserted surgically. Of the 109 patients with successful tube conversion, jejunal tip malposition occurred at follow-up in 18 (16.5%). Of these, four patients developed aspiration pneumonia (22%), which contributed to patient death in two. The frequency of jejunal tip malposition was 3.8% (one of 26 patients) for radiologically placed gastrostomy tubes and 20% (17 of 83 patients) for nonradiologically placed gastrostomy tubes (P ⴝ .07). Combined, 32% of gastrostomy tubes placed nonradiologically resulted in either procedural failure or eventual jejunal tip malposition, compared to 7.4% of radiologically placed gastrostomy tubes (P ⴝ .01). CONCLUSIONS: The frequency of procedural failure or eventual jejunal tip malposition with conversion of radiologically placed gastrostomy tubes to gastrojejunostomy tubes is significantly lower with radiologically placed gastrostomy tubes than with nonradiologically inserted gastrostomy tubes. J Vasc Interv Radiol 2010; 21:1031–1037
PERCUTANEOUS gastrostomy tubes are often necessary for patients with imFrom the Department of Radiology, Division of Vascular and Interventional Radiology, Duke University Medical Center, Box 3808, Durham, NC 27710 (C.Y.K., M.B.P., M.J.M., P.V.S., T.P.S.); and University Radiology, Knoxville, Tennessee (A.B.). Received April 2, 2009; final revision received October 28, 2009; accepted April 3, 2010. Address correspondence to C.Y.K.; E-mail: [email protected] None of the authors have identified a conflict of interest. © SIR, 2010 DOI: 10.1016/j.jvir.2010.04.003
paired neurologic or gastrointestinal status for the purposes of enteral alimentation and gastric decompression (1–3). However, aspiration pneumonia in the setting of esophageal reflux poses a considerable risk in patients with gastrostomy tubes. The benefit of gastrojejunostomy tube placement is a markedly decreased risk of enteral alimentation reflux by virtue of the jejunal tip being positioned beyond the ligament of Treitz (4). Radiologic conversion of a gastrostomy tube to a gastrojejunostomy tube involves removal of the gastrostomy
tube and insertion of a gastrojejunostomy tube through the preexisting tract, with subsequent negotiation of the jejunal portion of the tube through the pylorus into the duodenum and proximal jejunum by using fluoroscopic guidance. However, it has been suggested that an unfavorable angulation of the gastrostomy tube tract relative to the pylorus may result in failure to cannulate the duodenum (5,6). Furthermore, in the presence of such suboptimal tract angulation, the jejunal portion of the gastrojejunostomy tube may be prone to retrograde migration into the stomach.
1031
1032
•
Gastrostomy-to-gastrojejunostomy Tube Conversion
Accordingly, it has been our experience that some gastrojejunostomy tubes are susceptible to multiple episodes of retraction of the tip of the catheter back into the stomach, requiring multiple repeat procedures to reposition the tip back into the jejunum. Percutaneous gastrostomy tubes can be placed by using open or laparoscopic surgical, endoscopic, or radiologic techniques (7–11). During radiologic placement of gastrostomy tubes, a favorable angulation toward the pylorus is typically used. However, this angulation toward the pylorus can be very difficult to achieve with nonradiologic gastrostomy tube placement. The purpose of this retrospective study was to evaluate whether the method of gastrostomy tube placement affects the success of gastrostomy-to-gastrojejunostomy tube conversion and subsequent jejunal tip malposition with currently used large-bore gastrojejunostomy tubes.
MATERIALS AND METHODS This retrospective study was approved by the institutional review board, and a waiver of the informed consent requirement was obtained for this Health Insurance Portability and Accountability Act– compliant study. Patient Selection and Data Collection From January 2000 through December 2008, 130 patients underwent gastrostomy-to-gastrojejunostomy tube conversion at our institution. Six patients had a previously performed surgical gastrojejunostomy anastomosis and are not included here; therefore, the study cohort consists of 124 adult and pediatric patients. The mean patient age was 60 years (range, 13– 87 years; median, 61 years). There were 72 male and 52 female patients. The gastrostomy tubes were initially inserted radiologically in 27 patients, endoscopically in 75, with open surgery in 20, and laparoscopically in two. The most common primary indication for gastrostomy-to-gastrojejunostomy tube conversion was aspiration or risk of aspiration (n ⫽ 83), followed by gastroparesis (n ⫽ 24), pancreatitis (n ⫽ 6), leakage from the gastrostomy tube exit site (n ⫽ 6), gastric outlet obstruction (n ⫽ 3), gastric perforation (n ⫽ 1), and superior mesenteric artery syndrome (n ⫽ 1). Indications for
original gastrostomy tube insertion included respiratory failure with prolonged intubation (n ⫽ 44), failure to thrive (n ⫽ 28), neurologic deficit with impaired swallowing (n ⫽ 23), head and neck tumor (n ⫽ 9), gastric outlet obstruction (n ⫽ 9), esophageal perforation (n ⫽ 4), aspiration (n ⫽ 4), and severe dysphasia (n ⫽ 3). The clinical, radiologic, and surgical notes were reviewed for procedural success and gastrojejunostomy tube tip position at follow-up. Complications were assessed in accordance with the SIR Reporting Standards (12). Gastrostomy Tube–to– gastrojejunostomy Tube Conversion Technique All procedures were performed in an interventional radiology suite using fluoroscopy. Procedures were performed by an attending interventional radiologist or an interventional radiology fellow under the supervision of an attending radiologist. Conscious sedation with midolazam and fentanyl was used for all patients. The preexisting gastrostomy tube was removed over a guide wire, which was then manipulated into the pylorus and duodenum by using a 5-F angled catheter. If this was unsuccessful, multiple catheters and devices were used to cannulate the pylorus. These included modified catheters, vessel dilators, transjugular liver biopsy sets (Liver Access and Biopsy Set; Cook, Bloomington, Indiana), and transjugular intrahepatic portosystemic shunt sets (Rosch-Uchida Transjugular Liver Access Set, Cook) to increase the chances of successful small bowel catheterization. Once the guide wire was positioned in the proximal jejunum, the catheter was removed and a new 18-F MIC gastrojejunostomy tube (Medical Innovation, Draper, Utah) was inserted over the guide wire with the tip in the fourth portion of the duodenum or beyond the ligament of Treitz. Injection of a small amount of iodinated contrast medium was used to confirm appropriate jejunal tip position. The gastric port was then injected with iodinated contrast medium or air to confirm appropriate intragastric positioning of the gastric vent. The intragastric retention balloon was then inflated with saline, sometimes with a tiny amount of contrast medium. If the guide wire or gastrojejunostomy tube could not be ma-
July 2010
JVIR
nipulated through the pylorus into the duodenum despite multiple attempts, the procedure was aborted and a gastrostomy tube reinserted (Medical Innovation). The former was deemed a successful procedure, with the latter considered a failure of gastrostomy-togastrojejunostomy tube conversion. Radiologic Gastrostomy Tube Insertion Technique Briefly, patients were administered barium orally or via a nasogastric tube at least 12 hours prior to the planned procedure, in order to opacify the transverse colon. After gastric insufflation, a gastrostomy tube site was chosen below the costal margin, above the transverse colon, and to the left of midline. Two percutaneous T-fasteners were guided fluoroscopically into the stomach and deployed on either side of the planned gastrostomy tube incision. A needle was inserted into the stomach angled toward the pylorus, through which a guide wire was inserted. After serial dilation of the tract, an 18-F MIC gastrostomy tube was inserted. Endoscopic Gastrostomy Tube Insertion Technique After advancement of a gastroscope via the mouth into the stomach, the stomach was insufflated with air. Pressure was applied externally in the left upper quadrant two finger breadths below the left costal margin. As the light source was advanced to the area of maximal indentation internally, transillumination through the skin was seen externally. An 18-gauge needle was then passed though the anterior abdominal wall into the stomach under direct vision. A guide wire was passed into the stomach and then grasped with an endoscopic snare and pulled retrograde out the mouth. A gastrostomy tube was then advanced anterograde through the mouth, into the stomach, and out the anterior abdominal wall. Open Surgical Gastrostomy Tube Insertion Technique Open gastrostomy tube insertions were performed in a Stamm fashion. Briefly, a short upper midline incision was made and the peritoneal cavity entered. At a relatively avascular area in the mid-body of the anterior stomach,
Volume 21
Number 7
two silk purse-string stitches are placed. A 24-F gastrostomy tube was inserted through a separate stab incision in the left upper quadrant and into an incision located at the center of the pre-placed purse-string stitches, followed by balloon inflation and tightening of the purse-string stitches. The stomach wall adjacent to the tube insertion site was then sutured to the anterior abdominal wall to allow solid fusion to take place and to prevent leakage into the peritoneal cavity. Laparoscopic Gastrostomy Tube Insertion Technique A laparoscope was inserted via a 1-cm midline incision and the peritoneal cavity insufflated with air. A 1-cm incision was made in the left upper quadrant for placement of a 5-mm port. The stomach was insufflated with air via a nasogastric tube and then retracted toward the left upper quadrant port site. The stomach was then fixed to the anterior abdominal wall by using four Tfasteners or nylon sutures between the stomach and anterior abdominal wall under laparoscopic visualization. Once the stomach was secured to the abdominal wall, a needle was advanced into the stomach, through which a guide wire was inserted. In a Seldinger fashion, the wire was exchanged with a series of enlarging dilators. Once the tract was sufficiently dilated, a gastrostomy tube was inserted through a peel-away introducer sheath. Statistical Methods Data analysis was performed using software (SAS version 9; SAS Institute, Cary, North Carolina). Statistical analysis of technical failure and malposition rates was performed and correlated with the method of original gastrostomy placement, indication for initial gastrostomy tube insertion, indication for gastrostomy-to-gastrojejunostomy tube conversion, and length of time between gastrostomy insertion and conversion by using the Fisher exact test. A P value of less than or equal to .05 was considered statistically significant.
RESULTS Of the 124 patients referred for gastrostomy-to-gastrojejunostomy tube conversion with a surgically nonaltered
Kim et al
•
1033
Table 1 Frequency of Procedural Failures and Jejunal Catheter Tip Malposition according to Method of Gastrostomy Placement in Stomachs without Surgical Alteration Gastrostomy Method Radiologic Endoscopic Open surgical Laparoscopic Nonradiologic Total
No. of Patients
Procedural Failure
Jejunal Tip Malposition*
Failure or Malposition
27 75 20 2 97 124
1 (3.7) 12 (16) 2 (10) 0 (0) 14 (14) 15 (12.1)
1 (3.8) 14 (22) 3 (17) 0 (0) 17(20) 18 (16.5)
2 (7.4) 26 (35) 5 (25) 0 (0) 31 (32) 33 (26.6)
Note.—Numbers in parentheses are percentages. * Number of patients for percentage calculation equals number of patients minus patients with procedural failure.
stomach, 15 (12.1%) had procedural failures due to inability to manipulate a guide wire into the duodenum (Table 1). One of the 27 radiologically placed gastrostomy tubes (3.7%) resulted in failed conversion to a gastrojejunostomy tube due to inability to cannulate the pylorus, whereas 14 of the 97 nonradiologically placed gastrostomy tubes (14%) resulted in conversion failure (P ⫽ .19); (Figs 1, 2). Follow-up data were available for an average of 136 days after gastrostomy tube–to– gastrojejunostomy tube conversion (median, 63 days; range, 1–1,300 days). Endpoints for follow-up included the following: lost to follow-up (n ⫽ 39), death (n ⫽ 30), gastrojejunostomy tube removal because no longer needed (n ⫽ 23), currently in place (n ⫽ 5), infection (n ⫽ 3), conversion to a gastrostomy tube due to recurrent malposition (n ⫽ 3), gastrojejunostomy tube fell out (n ⫽ 2), de novo gastrojejunostomy tube placement at new site due to recurrent malposition (n ⫽ 1), and gastrojejunostomy tube removal due to exit-site excoriation (n ⫽ 1). Of the 109 patients with successful gastrostomy tube–to–gastrojejunostomy tube conversion, 18 had eventual retrograde malposition of the jejunal catheter into the stomach. The average interval between gastrostomy tube conversion and jejunal catheter malposition was 116 days (median, 47 days; range, 5–1,059 days). The average number of subsequent jejunal catheter repositioning procedures in each of these 18 patients was 1.8 (median, 1; range, 1–10). Correlation with the method of initial gastrostomy tube placement revealed that one of the 26 (3.8%) radiologically placed gastrostomy tubes that were successfully con-
verted to gastrojejunostomy tubes had eventual jejunal tip malposition compared to 17 of the 83 (20%) nonradiologically placed gastrostomy tubes that were successfully converted to gastrojejunostomy tubes (P ⫽ .07). The final disposition of the 18 malpositioned gastrojejunostomy tubes was as follows: three were removed because they were no longer needed, two were converted to a gastrostomy tube, one was replaced with a new gastrojejunostomy tube inserted via a de novo puncture, and one was removed and replaced with a laparoscopically inserted jejunostomy tube. In the remaining 11 patients, the patient either died (n ⫽ 6) or was lost to follow-up (n ⫽ 4), or the gastrojejunostomy tube is currently in use (n ⫽ 1). Four of the 18 patients with malpositioned gastrojejunostomy tubes had aspiration pneumonia that was directly attributed to continued tube feeding through a gastrojejunostomy tube, which had migrated proximally into the stomach or esophagus (Table 2). In three of these cases, the tip of the gastrojejunostomy tube was found to be in the esophagus (Fig 3). In two of the patients with gastrojejunostomy tube tip malposition, the clinical team had deemed that the tube malposition-related aspiration pneumonia contributed to the death of the patient. One patient was admitted four times due to aspiration pneumonia associated with recurrent gastrojejunostomy tube tip malposition. Overall, 7.4% of radiologically inserted gastrostomy tubes resulted in gastrostomy tube conversion failure or eventual malposition, compared to 32% of nonradiologically placed gastrostomy tubes (P ⫽ .01). No statistically signifi-
1034
•
July 2010
Gastrostomy-to-gastrojejunostomy Tube Conversion
Figure 1. Favorable tract angulation with radiologic gastrostomy tube insertion. (a) Radiograph obtained after injection of iodinated contrast medium through the radiologically inserted gastrostomy tube (arrows) in a 64-year-old woman who required enteral alimentation due to anoxic brain injury. Note the gastric insertion site at the proximal gastric antrum with orientation of the tract toward the pylorus (*). (b) Axial computed tomographic (CT) scan in a 76-year-old woman after radiologic gastrostomy tube insertion due to poor functional status and malnutrition. Notice the transabdominal tract angulation (arrows) at approximately 40° in relation to the skin toward the pylorus (*).
Figure 2. Unfavorable tract angulation of a gastrostomy tube originally inserted endoscopically. (a) Radiograph obtained after injection of a small amount of iodinated contrast medium through the gastrostomy tube in a 24-year-old man in need of enteral nutrition due to traumatic brain injury and poor functional status. The gastrostomy tube (arrows) had been previously exchanged for a balloon-tip gastrostomy tube radiologically. Note the gastric insertion site at the mid-gastric body, with tract orientation toward the gastric fundus (*). (b) CT scan in an 81-year-old man in need of enteral nutrition due to traumatic brain injury and poor functional status. The gastrostomy tube (arrows) had also been previously exchanged for a balloon-tip gastrostomy tube radiologically. The CT scan demonstrates tract angulation toward the patient’s left side, away from the pylorus. Due to the tract angulation and direction, this gastrostomy tube will likely be difficult to convert to a gastrojejunostomy tube and will likely be prone to retrograde tip malposition into the stomach.
cant differences were found between the nonradiologically inserted gastrostomy tubes. In addition, patients who
were referred for gastrostomy-to-gastrojejunostomy tube conversion with the indication of gastric outlet obstruc-
JVIR
tion or gastroparesis were significantly more likely to experience a technical failure of conversion (25% vs 8%, P ⫽ .05), although there was no significant difference in eventual malposition rates. The indication for initial gastrostomy tube insertion and length of dwell time of the gastrostomy tube before conversion were not significantly correlated with technical success or eventual tip malposition. After gastrostomy tube conversion, exit site infection occurred in three patients. Subsequent catheter removal was performed in all three patients to help resolve the infection. However, in two of these patients, pericatheter leakage of gastric contents and signs and symptoms of exit site infection were present before the gastrostomy tube conversion procedure and, in fact, were the indications for gastrostomy tube conversion. In one of these patients, the catheter was removed on postprocedure day 1, with subsequent de novo gastrostomy tube insertion via laparoscopic surgery. The other patient’s catheter was removed on postprocedure day 8, and a nasogastric tube and nasojejunal feeding tube were inserted to address the patient’s enteral requirements until the patient’s death 3 days later. The third patient’s gastrojejunostomy tube was removed 13 days after gastrostomy tube conversion due to new cellulitis around the gastrojejunostomy tube exit site that developed after the conversion procedure. However, the patient’s gastroparesis had resolved in the interim, and therefore no additional intervention was performed after catheter removal. The only procedural complication was this single exit site infection. No other complications related to gastrostomy tube conversion were identified.
DISCUSSION Both gastric and jejunal tube feeding have proven to be effective methods for providing nutrition to patients who cannot tolerate oral alimentation (13). Whereas jejunal feeding is typically infused in a continuous manner, gastric feeding may have a physiologic benefit and has the option of bolus infusion, which is particularly convenient for the non-bed-ridden patient. At our institution, gastrostomy tube insertion is frequently the initial intervention of choice for these reasons and also because gastrostomy tube insertion is technically
Volume 21
Number 7
Kim et al
Table 2 Complications Associated with Gastrojejunostomy Tube Tip Malposition Event
No. of Patients (n ⫽ 18)
No. of Episodes (n ⫽ 33)
Gastrojejunostomy tube tip in esophagus Aspiration pneumonia Death related to aspiration pneumonia
4 (22) 4 (22) 2 (11)
11 (33) 7 (21) NA
Note.—Numbers in parentheses are percentages. NA ⫽ not applicable.
easier, faster, and less expensive than gastrojejunostomy tube insertion and has fewer episodes of clogging (7,11). However, in the setting of aspiration risk, gastric dysmotility, or pancreatitis, primary gastrojejunostomy tube insertion is indicated. If these conditions occur or are discovered after gastrostomy tube insertion, then gastrostomy-to-gastrojejunostomy tube conversion is indicated. At our institution, gastrostomy tubes are initially inserted endoscopically or radiologically on the basis of referring physician preference or availability. Open gastrostomy tube insertion typically occurs only when a patient undergoes major gastrointestinal surgery, where nontransient enteral feeding requirements are anticipated. During endoscopic gastrostomy tube placement (14), the orientation of the gastrostomy tube tract can be difficult to dictate. The tract, established by means of initial needle insertion, tends to be perpendicular to the skin and stomach due to safety concerns for injury to nonvisible intraabdominal structures, such as the colon, and is sometimes angled toward the gastroesophageal junction. Depending on the anatomy of the stomach, this can create an acute angle for a guide wire or gastrojejunostomy tube to reach the duodenum. Similarly, during open surgical or laparoscopic gastrostomy tube placement (15,16), the tract is also often created relatively perpendicular or randomly in terms of angulation with the skin and stomach. In fact, an angulated tract is often avoided during open surgical gastrostomy tube insertion to allow an adequate fascial edge for closure. Because gastrostomy tube–to– gastrojejunostomy tube conversion is not typically performed by most surgeons or endoscopists, the issue of tract angulation is likely not taken into account during gastrostomy tube placement. Although it is possible to attempt to create
the tract in a favorable orientation toward the pylorus during surgical gastrostomy placement, the final geometric orientation of the tract is difficult to predict because the tract is created under nonanatomic conditions. During open surgery, the gastrostomy incision is made while the stomach is exteriorized through the primary incision and then a separate incision made through the skin for the planned exit site. Therefore, the tract angulation may change considerably once the stomach returns to its natural, in vivo configuration. Similarly, during laparoscopic gastrostomy tube placement, the gastrostomy incision and skin incision are made during peritoneal insufflation, which alters the normal anatomic relationship of the stomach with regard to the planned exit site. Conversely, during radiologic placement of gastrostomy tubes creation of a favorable tract angulation is straightforward to perform. Because the tract is created by needle insertion through the skin into the stomach during real-time fluoroscopic visualization of the stomach in its normal in vivo configuration, an exaggerated angle can be created toward the pylorus, in effect, as much as desired. Furthermore, because the gastric puncture is made with the anterior stomach wall already fixed to the anterior abdominal wall, this angulation will change minimally after completion of the procedure. Finally, because gastrostomy tube conversion is most commonly performed by interventional radiologists at our institution, an appropriate tract orientation is routinely used for gastrostomy tube placement in case later gastrojejunostomy tube conversion is required. On the basis of our experience and that of others (5,6), we believe that suboptimal tract orientation created at the time of initial gastrostomy tube insertion is the primary cause of procedural conversion failures. An unfavorable
•
1035
tract orientation in relation to the gastric lumen and pylorus can make it very difficult or impossible to manipulate a guide wire and catheter into the duodenum due to the poor mechanical vector and tendency for buckling into the fundus of the stomach (Figs 1, 2). In addition, the site of original puncture of the stomach for gastrostomy and anatomic configuration of the stomach and pylorus may also play a role. Although there was a tendency for greater success with conversion of radiologically placed gastrostomy tubes (96.7%) as compared to nonradiologically placed gastrostomy tubes (85.6%), this did not reach statistical significance. However, not surprisingly, gastrostomy conversion procedures with gastroparesis or gastric outlet obstruction as the indication were significantly more likely to result in technical failure. No other definite predictors of success or failure were identified. An unfavorable tract angle is also presumed to be the reason for subsequent retrograde malposition of the jejunal portion of the gastrojejunostomy tube into the stomach. In our series, 16.5% of patients experienced retrograde malposition of the jejunal tip into the stomach and, of these, 22% resulted in the tip in the esophagus, which creates an extremely high aspiration risk even in patients without gastroesophageal reflux (Fig 3). In two of the patients, the associated aspiration pneumonia contributed to their deaths. Therefore, it is important for the operator to identify those patients at risk for gastrojejunostomy tube tip malposition and consider alternate procedures. It is also important for the patients’ physicians to be aware of this substantial risk for malposition with resultant aspiration, perhaps warranting more frequent monitoring with abdominal radiographs to assess gastrojejunostomy tube position in the setting of feeding intolerance. Moreover, repositioning of malpositioned gastrojejunostomy tubes is costly and incurs additional radiation dose to the operator and patient. In these patients with recurrent episodes of tip migration, primary jejunostomy tube insertion may be indicated (17). Alternatively, removal of the gastrostomy tube and creation of a new puncture tract with a more favorable angulation toward the pylorus could be performed for de novo gastrojejunostomy tube insertion. It should be kept in mind that additional possible contribu-
1036
•
Gastrostomy-to-gastrojejunostomy Tube Conversion
Figure 3. Images in a 68-year-old man with locked-in syndrome who required enteral nutrition due to poor functional status. Gastrostomy tube–to– gastrojejunostomy tube conversion was performed due to episodes of aspiration. (a) Fluoroscopic spot image obtained after successful gastrostomy tube conversion after injection of iodinated contrast medium through the jejunostomy and gastric port. Opacification of the proximal jejunum (arrowheads) helps confirm the intrajejunal location of the tip (arrow). (b) Two months later, chest radiograph demonstrates the gastrojejunostomy tube to be coiled in the gastric fundus, with the jejunal tip extending into the proximal esophagus (arrows). This patient had aspiration pneumonia secondary to enteral feeding through this tube. On some of this patient’s chest radiographs, the malpositioned gastrojejunostomy tube was mistaken for a nasogastric tube.
tors to jejunal catheter malposition may include excessive peristalsis, reverse peristalsis, vomiting, gastric anatomy, and anatomic location of gastric puncture. On the basis of our review of the literature, only one article has compared success rates of gastrostomy tube–to– gastrojejunostomy tube conversion on the basis of the method of original gastrostomy tube placement (5). In that study, 38 patients had surgical gastrostomy tube placement (with 18% procedural failure) and 18 had endoscopic gastrostomy tube placement (with 22% procedural failure); only seven had radiologic gastrostomy tube placement (with no procedural failures). No statistical analysis was specified, likely because the number of patients with radiologic gastrostomy tube placement was inadequate to achieve statistical significance. In terms of evaluation of eventual jejunal tip malposition (n ⫽ 8), the mean length of follow-up was only 9 weeks (range, 1–16 weeks). Furthermore, their analysis was based on small-bore (12–16-F) catheters, some or all of which were single-lumen transgastric jejunal feeding catheters, whereas the ones analyzed in our study are the more commonly used large-bore 18-F
gastrojejunostomy tubes with both a gastric and jejunal port. These largerbore catheters tend to be stiffer, which has an unknown effect on the likelihood for buckling into the stomach and subsequent jejunal tip malposition. The primary limitation of this study is its retrospective nature, which prohibited analysis of a number of parameters that may help predict gastrostomy tube conversion failure and eventual gastrojejunostomy tube tip malposition. As mentioned, the authors and others believe that the gastrostomy tube tract orientation has a substantial effect on the outcome of gastrostomy tube conversion (5,6). However, a formal analysis of tract orientation was not possible due to a lack of assessment at conversion. Unfortunately, a very small minority of the patients underwent CT after gastrostomy tube insertion and before conversion to a gastrojejunostomy tube, which would be helpful for objectively assessing tract orientation and measuring tract angles. In addition, the gastric anatomy and actual location of gastric puncture could not be retrospectively assessed due to the variability in abdominal imaging acquisition and variable presence of adequate intraluminal air or contrast
July 2010
JVIR
medium to delineate gastric anatomy. Furthermore, a subjective assessment by the interventional radiologist performing the procedure was inconsistently documented, which may have also provided potentially predictive information. In our series, the interventional radiologists went to great lengths to successfully convert gastrostomy tubes to gastrojejunostomy tubes, using a wide array of interventional tools and prolonged fluoroscopy times, for example, as high as 42 minutes. Unfortunately, procedural details of these additional measures and devices used for gastrostomy tube conversion were inconsistently reported and fluoroscopy times were not documented for most procedures. This information may have allowed for a cost analysis that could potentially affect decision making. Considering the median postgastrostomy tube conversion follow-up of 63 days and median time to jejunal tip malposition of 47 days, 48 patients (44%) had follow-up data for fewer than 47 days. Therefore, the malposition rates in our study are likely underestimated. Although this study represents the largest number of gastrostomy tube–to– gastrojejunostomy tube conversions in the literature, the number of procedures is somewhat limited. With a larger number of patients, both the conversion success rate and jejunal tip malposition rate may have achieved statistical significance. In summary, the results of this study demonstrated that gastrostomy tubes inserted radiologically have improved outcomes when converted to gastrojejunostomy tubes, in terms of procedural failure rates and eventual malposition rates, when compared to nonradiologically inserted gastrostomy tubes. Our results suggest that patients with a high probability for gastrostomy tube conversion may benefit from radiologic placement of the original gastrostomy tube with intentional angulation toward the pylorus. When faced with a patient with a nonradiologically placed gastrostomy tube in need of a gastrojejunostomy tube, a careful assessment of the gastrostomy tube tract orientation in relation to the pylorus is likely warranted. In cases of poor tract angulation or difficulty cannulating the pylorus, primary jejunostomy tube insertion or a de novo gastrojejunostomy tube access with a more favorable orientation should be considered in light of the 26.6% fre-
Volume 21
Number 7
quency of procedural failure or eventual jejunal tip malposition for all procedures. Although overall procedural success rates are high with all methods of gastrostomy tube placement, a dangerous scenario is subsequent occult retrograde catheter tip malposition, which resulted in a 22% frequency of aspiration pneumonia in this largely debilitated patient population. Therefore, operator and referring clinician awareness of this not-infrequent complication should be emphasized. In addition, gastrostomy tube insertion with nonradiologic techniques may benefit from procedural modification to minimize procedural failure and jejunal tip malposition when subsequent conversion to a gastrojejunostomy tube is required. References 1. O’Keeffe F, Carrasco CH, Charnsangavej C, Richli WR, Wallace S, Freedman RS. Percutaneous drainage and feeding gastrostomies in 100 patients. Radiology 1989; 172:341–343. 2. Halkier BK, Ho CS, Yee AC. Percutaneous feeding gastrostomy with the Seldinger technique: review of 252 patients. Radiology 1989; 171:359 –362. 3. van Sonnenberg E, Wittich GR, Cabrera OA, et al. Percutaneous gastrostomy and gastroenterostomy. II. Clinical experience. AJR Am J Roentgenol 1986; 146:581–586.
Kim et al
4. Gustke RF, Varma RR, Soergel KH. Gastric reflux during perfusion of the proximal small bowel. Gastroenterology 1970; 59:890 – 895. 5. Lu DS, Mueller PR, Lee MJ, Dawson SL, Hahn PF, Brountzos E. Gastrostomy conversion to transgastric jejunostomy: technical problems, causes of failure, and proposed solutions in 63 patients. Radiology 1993; 187:679 – 683. 6. McLean GK, Rombeau JL, Caldwell MD, Ring EJ, Freiman DB. Transgastrostomy jejunal intubation for enteric alimentation. AJR Am J Roentgenol 1982; 139:1129 –1133. 7. Barkmeier JM, Trerotola SO, Wiebke EA, et al. Percutaneous radiologic, surgical endoscopic, and percutaneous endoscopic gastrostomy/gastrojejunostomy: comparative study and cost analysis. Cardiovasc Intervent Radiol 1998; 21:324 –328. 8. Silas AM, Pearce LF, Lestina LS, Grove MR, et al. Percutaneous radiologic gastrostomy versus percutaneous endoscopic gastrostomy: a comparison of indications, complications and outcomes in 370 patients. Eur J Radiol 2005; 56:84 –90. 9. Consentini EP, Sautner T, Gnant M, Winkelbauer F, Teleky B, Jakesz R. Outcomes of surgical, percutaneous endoscopic, and percutaneous radiologic gastrostomies. Arch Surg 1998; 133: 1076 –1083. 10. Wollman B, D’Agostino HB, WalusWigle JR, Easter DW, Beale A. Ra-
11.
12.
13. 14.
15.
16.
17.
•
1037
diologic, endoscopic, and surgical gastrostomy: an institutional evaluation and meta-analysis of the literature. Radiology 1995; 197:699 –704. Hoffer EK, Cosgrove JM, Levin DQ, Herskowitz MM, Sclafani SJ. Radiologic gastrojejunostomy and percutaneous endoscopic gastrostomy: a prospective, randomized comparison. J Vasc Interv Radiol 1999; 10:413– 420. Sacks D, McClenny TE, Cardella JF, Lewis CA. Society of Interventional Radiology clinical practice guidelines. J Vasc Interv Radiol 2003; 14: S199 –S202. Jabbar A, McClave SA. Pre-pyloric versus post-pyloric feeding. Clin Nutr 2005; 24:719 –726. Safadi BY, Marks Jm, Ponsky JL. Percutaneous endoscopic gastrostomy. Gastrointest Endosc Clin N Am 1998; 8:551–568. Apelgren KN, Zambos J. Is percutaneous better than open gastrostomy? A clinical study in one surgical department. Am Surg 1989; 55:596 – 600. Murayama KM, Johnson TJ, Thompson JS. Laparoscopic gastrostomy and jejunostomy are safe and effective for obtaining enteral access. Am J Surg 1996; 172:591–594. Tapia J, Murguia R, Garcia G, de los Monteros PE, Onate E. Jejunostomy: techniques, indications, and complications. World J Surg 1999; 23:596 – 602.