Enteral feeding through gastrojejunal (GJ) tubes is an established method of nutrition for patients with feeding difficulty who do not tolerate intragastric feedings through gastrostomy (G) tubes. Patients requiring jejunal feeding are often high-risk surgical candidates, so endoscopically and/or fluoroscopically placed tubes offer a reasonable alternative. Literature about the long-term outcome, safety and complications of different GJ tubes and placement methods remains lacking in children. In the present study we describe our pediatric center's experience with use of GJ tubes by pediatric gastroenterologists for 10 years.
This was a retrospective chart review for GJ tube placement procedures in all patients followed by the Pediatric Gastroenterology Division at the University of Iowa from August 1, 1999 to July 31, 2009. Patients were identified through a procedure database and billing records. The institutional review board of University of Iowa approved the study.
For the purposes of this study, we defined an initial GJ tube placement as the first successful GJ tube placement during the study period. All subsequent GJ tubes placed were considered replacements. Beginning of follow-up was defined as the time of initial successful GJ tube placement. End of follow-up time was defined as the time of surgical intervention (eg, fundoplication, surgical jejunostomy) or (if no surgical intervention performed) the last documented clinic or phone encounter while using a nonsurgical feeding (GJ or G) tube, patient death, or end of study period. Replacement indications were classified as accidental dislodgement, tube obstruction, tube coiling in the stomach (verified radiologically), broken tube component, hole in the tube, balloon rupture, and the occurrence of intussusception or bowel tenting. Bowel tenting was diagnosed based on radiologic evidence of tenting (with contrast imaging) in the setting of abdominal pain and/or gastrointestinal bleeding. Patients with and without established gastrocutaneous tracts at the time of the first GJ placement were included. Indications for tube replacement and type of sedation used were analyzed and compared between tube types. Early tube replacement was defined as occurring within 14 days of tube placement.
Tubes and Placement Methods
The most common tube types used during the study were low-profile MIC-Key transgastric jejunal (LP MIC-Key GJ) feeding tube (Kimberly-Clark, Roswell, GA) and coaxial percutanous endoscopic gastrostomy-percutanous endoscopic jejunostomy (PEG-PEJ) feeding tube system (Cook Medical, Winston-Salem, NC) (Fig. 1). Three placement methods were used in patients with established gastrocutaneous tracts:
- Method 1: Guidewire placement with fluoroscopy with subsequent GJ tube placement without use of endoscopy
- Method 2: Guidewire placement during gastroduodenoscopy through the gastrostomy site followed by GJ tube placement
- Method 3: Tube placement during esophagogastroduodenoscopy (EGD)
Fluoroscopic placement (method 1) was typically used only for replacement of an existing, well-positioned GJ tube that was malfunctioning. In this case, a guidewire is placed into the duodenum via the existing tube, following which the old tube is removed and a new one is placed over the guidewire. Blind (nonendoscopic) guidewire placement through the gastrocutaneous tract aiming for duodenal intubation is also possible (used by radiologists) but is rarely used in our practice. Placement during gastroduodenoscopy (method 2) combines endoscopy and fluoroscopy. A pediatric endoscope (outer diameter 5.5–5.9 mm) is placed through the gastrocutaneous tract into the distal duodenum. A guidewire is advanced through the endoscope under direct vision and its position is verified by fluoroscopy. The endoscope is then retrieved and a GJ tube is advanced over the guidewire into position, using brief bursts of fluoroscopy to confirm proper advancement. GJ tube placement during EGD (method 3) was seldom used. In this technique, the endoscope is introduced through the mouth into the stomach, to assist with guidewire or GJ tube advancement through the pylorus into the small bowel.
In the absence of an established gastrocutaneous tract, primary placement of a coaxial PEG-PEJ feeding tube system can be performed. A 20-F percutanous endoscopic gastrostomy (PEG) tube is placed using the pull technique during an EGD (1). A guidewire is then advanced through the PEG tube into the stomach. Under fluoroscopic guidance or with the help of endoscopy, the guidewire is advanced into the small bowel and a 9-F jejunal feeding tube with a special dual port connection is passed over the guidewire into the small bowel (Fig. 1).
Continuous data were reported as means, medians, standard deviations, and ranges and were compared by an unpaired Student t test or the Wilcoxon signed rank test. Categorical variables were described as percentages and proportions and were compared by a χ 2 test or Fisher exact test (depending on the expected cell count).
Univariate survival analysis for time to tube replacement (tube survival) was performed according to the Kaplan-Meier method. Survival time was defined as the time interval from tube placement to replacement with another GJ tube, a G tube, surgical intervention, patient death, or end of the follow-up. Censored events were defined as those when a GJ tube remained functional at time of removal (eg, at switch to G tube, surgical intervention, loss of follow-up). A 2-tailed log-rank test was applied to determine statistically significant differences. P < 0.05 was considered significant. Statistical analysis was performed with SAS 9 statistical software (SAS Institute, Cary, NC) and GraphPad Prism 4 for Windows (Graphpad Software Inc, San Diego, CA).
The study included 33 patients with a median age of 2.5 years (range 7.1 months–21.6 years). Cohort characteristics are shown in Table 1. The main indication for GJ tube placement (67%) was to provide enteral nutrition access in patients with intragastric feeding intolerance secondary to severe gastroesophageal reflux disease (GERD). The most common underlying diagnosis was neurologic impairment (76%), followed by congenital heart disease and chronic lung disease with a history of prematurity. Some patients had multiple underlying diagnoses and/or indications for GJ tube placement. The majority of patients in this cohort (N = 31, 94%) had mature gastrocutaneous tracts at the beginning of the study. Two patients in our series did not have gastrostomy tubes placed because they failed NG feeding but tolerated either transpyloric or nasojejunal tube feeds and therefore underwent direct GJ tube placement.
During the study period, a total of 164 placement attempts were made, of which 160 were successful (Fig. 2). Four failed attempts were noted, with an overall success rate of 97.6%. Tubes successfully placed consisted of 114 LP MIC-Key GJ tubes, 38 coaxial PEG-PEJ tubes, and 8 other nonprofile GJ tubes. The category of nonprofile GJ tubes included 2 initial placements and 6 replacements. There were a total of 33 initial successful placements and 127 replacements as previously defined. The study cohort included 2 patients without an existing gastrostomy at the time of initial placement who underwent successful coaxial PEG-PEJ tube placement. Procedure characteristics are summarized in Table 2. The mean number of GJ tube placements was 4.9 per patient (range 1–20) with a mean follow-up of 26.8 months (range 0.4–115.3 months). A change in practice was noted after the year 2005, with a sharp shift from use of coaxial PEG-PEJ tubes to LP MIC-Key GJ tubes (Fig. 3).
The indications for tube replacements are summarized in Table 3 and divided according to tube type. The main replacement indications included accidental tube dislodgement, tube obstruction (presumably with formula and/or medications), coiling back into the stomach, balloon rupture, and broken tube component. Comparison between the 2 main tube types showed that certain tube-related complications were more commonly encountered with the use of a certain tube type. This included a statistically higher rate of accidental tube dislodgement with coaxial PEG-PEJ tubes and exclusive occurrence of balloon ruptures with LP MIC-Key GJ tubes. Twenty-six tube replacements were early events (occurred within 14 days of placement). When classified by time to tube replacement, early events included coiling back into the stomach (P = 0.0017) and bowel tenting (P < 0.0001).
Most procedures (85.4%) required no sedation and were performed via a mature gastrocutaneous tract. Conscious sedation, typically in the form of intravenous meperidine and midazolam, was used in 13% of all cases, including 1 patient without an established gastrocutaneous tract. General anesthesia was used in 2 patients (1.2%) for coaxial PEG-PEJ tube placement, 1 of which did not have an established gastrocutaneous tract.
Stratifying by GJ tube type, successful LP MIC-Key GJ tube placement (N = 114) was performed without sedation in 104 cases (91.2%) and with conscious sedation in 8 procedures (7%). Clear data on sedation type were not available on 2 LP MIC-Key GJ tube placements (1.8%). In the coaxial PEG-PEJ tube group, 24 procedures (63.2%) required no sedation, 12 procedures (31.6%) required conscious sedation, and 2 procedures (5.3%) were done under general anesthesia. Comparing use of any sedation between the 2 major tube types placed showed statistically lower sedation use in the LP MIC-Key GJ group compared with the coaxial PEG-PEJ tube group (7.1% vs 36.8%, P < 0.0001).
Time to tube change was evaluated by the Kaplan-Meier analysis (Fig. 4), which included all tubes placed. Log-rank statistics showed a significantly longer time to tube change in the coaxial PEG-PEJ tube group (P = 0.04). To assess mean tube survival, the subset of tubes actively replaced because of malfunction or other complications was analyzed separately, whereas functional tubes removed (such as elective switch to G tube, surgical intervention, or loss of follow-up) were excluded. This analysis showed a mean tube survival in LP MIC-Key GJ and coaxial PEG-PEJ tube groups of 91 days (SD 107, range 1–522) and 177 days (SD 229, range 1–935), respectively (P = 0.06).
Four patients (12%) were lost to follow-up and their last GJ tube placement procedure was considered the end of follow-up. Long-term outcomes (Table 2) at the end of follow-up were as follows: 39% remained on jejunal feeds through a GJ tube, 30% were successfully switched to gastrostomy tube or oral feeds, 15% underwent surgical intervention (either Nissen fundoplication or surgical jejunostomy) and 15% were deceased (none thought related to the GJ tube use or placement).
Common indications for GJ tube feeds in children include GERD with high aspiration risk, intractable vomiting, failed Nissen fundoplication, and gastroparesis (2,3). The most frequent underlying diagnoses cited in the pediatric literature were neurologic disorders (3,4). Our cohort had similar characteristics, with most having GERD and/or documented aspiration associated with an underlying neurologic disorder. We believe that our cohort is a reasonable reflection of the patient population often seen in pediatric gastroenterology practice in tertiary care centers.
As mentioned, several methods exist to place GJ tubes in patients with established gastrocutaneous tracts (5). We primarily used methods 1 and 2 (guidewire placement with fluoroscopy or guidewire placement during gastroduodenoscopy through the gastrostomy site) because these circumvent the need for endoscope passage through the oropharynx, and therefore allow direct, rapid placement of the feeding tube while eliminating the need for sedation. Avoiding sedation was a major advantage in this high-risk patient cohort.
The use of fluoroscopic assistance may suggest to some that GJ tube placement should be preferentially performed by radiologists. Our data show that GJ tube placement by a pediatric gastroenterologist can be highly successful (success rate of 97.6%). Use of an endoscope allows much less radiation exposure, compared with the blind fluoroscopic approach, because guidewire placement into the small bowel is done under direct vision; however to our knowledge, such comparative data have not been published. It is important to emphasize that before embarking on GJ tube placement by pediatric gastroenterologists, proper training in radiation safety is mandatory for the well-being of patients and staff.
It is important to understand and discuss all aspects of GJ tube–associated complications with families. Major complications, including bowel perforation and intussusception, are rare (3,6). Our study included only 1 patient with intussusception, which we previously described (6). This complication was thought to be induced by mucosal injury from the tip of the feeding tube tenting the bowel at or near the ligament of Treitz. Minor complications occur more frequently and include tube obstruction, dislodgement, leakage, and migration (2,5). This was in agreement with our study findings, with minor complications leading to most tube replacements. Appropriate caregiver education, training, and nursing facility support are crucial to the maintenance and longevity of GJ tubes. Our study showed that 41.3% of LP MIC-Key GJ and coaxial PEG-PEJ tube replacements were because of preventable causes (obstruction and accidental pulling). Certain complications were more likely encountered within 2 weeks of placement (tube coiling back into the stomach and bowel tenting). Careful tube positioning during placement and proper instruction afterwards may help reduce replacement frequency from potentially avoidable events. To avoid obstruction, frequent jejunal port flushing with water is recommended, especially before and after formula and medication infusion, but adherence to this may not be universal. An alternative to manual flushing is the use of a programmable infusion pump that automatically flushes the tube with water at scheduled times, avoiding reliance on caregivers. Some centers have advocated for scheduled GJ tube replacements. In one study, scheduled GJ tube replacement every 4 months was claimed to prevent some mechanical tube-related complications such as obstruction (5); however, the optimal time for scheduled tube replacement is not known. Our data showed a mean tube survival of LP MIC-Key GJ tubes to be 91 days, and coaxial PEG-PEJ tubes to be 177 days, but the range was extremely wide. We presently keep functional GJ tubes in place for about 1 year or until signs of tube deterioration become apparent.
Some physicians may view GJ tube feedings as a temporary solution during periods of intragastric feeding intolerance and expect patients to eventually require a more permanent surgical intervention for long-term enteral nutrition (5,7,8). Our study showed that 39% of patients remained dependent on jejunal feeds through a GJ tube with another 15% proceeding to surgery. The decision to consider prolonged GJ tube use versus surgical intervention is likely influenced by several factors (available expertise, surgical risk and family preferences) and could not be assessed in the present study because of its retrospective nature. The study also showed that an appreciable subset (30% of patients) was successfully switched back to G tube or oral feeds (particularly in patients with temporary underlying conditions such as acute pancreatitis and superior mesenteric artery syndrome). It was not surprising that 15% of patients were deceased by the end of follow-up because the underlying diagnoses in this cohort make them a high-risk population with shorter life expectancy. Our study showed that long-term jejunal feeding via GJ tubes is possible and safe, but complications and costs associated with tube use and maintenance should be kept in mind. Mechanical problems pose a drawback to long-term use, leading to repeated procedures, radiation exposure, and increased health care costs.
No published literature exists as to the present practice or attitudes of pediatric gastroenterologists toward use of GJ tubes in children. Lack of training in this procedure during fellowship, center-specific procedural interests, access to fluoroscopy, and concerns about reimbursement and radiation exposure may be limitations to the use of GJ tubes among pediatric gastroenterologists. Further research is needed to understand attitudes, perceptions, and interest in acquiring GJ tube procedural placement skills among pediatric gastroenterologists.