Center for Motility and Functional Gastrointestinal Disorders, Children's Hospital Boston, Boston, MA, USA.
Received 11 May, 2010
Accepted 7 November, 2010
Address correspondence and reprint requests to Rachel Rosen, MD, MPH, Division of Gastroenterology and Nutrition, Children's Hospital Boston, 300 Longwood Ave, Boston, MA 02115 (e-mail: email@example.com).
This work was supported in part by NIDDK grants 1K23DK073713 and the Children's Hospital Boston Career Development Award.
The authors report no conflicts of interest.
Objectives: Transpyloric feeding has been proposed as an alternative method for controlling gastroesophageal reflux, but there have been no pediatric studies to determine how transpyloric feeding affects reflux burden. The aim of the present study was to determine the reflux burden in patients receiving transpyloric feeds.
Patients and Methods: We reviewed the multichannel intraluminal impedance tracings of patients who had persistent symptoms and were fed transpylorically during the multichannel intraluminal impedance recording. We compared the reflux profiles during feed and nonfeed periods. We also compared the number of reflux-related hospitalizations at Children's Hospital Boston in the year before and the year after the initiation of transpyloric feeds.
Results: The mean number of reflux events per hour was 1.4 ± 1.3 and 0.8 ± 1.1 during feed and nonfeed periods, respectively (P = 0.06). There was no significant difference in the percentage of time that boluses were present in the esophagus during feed periods (1.0% ± 1.4%) compared with the nonfeed periods (0.6% ± 1.1%, P = 0.5). There was no significant difference in the mean number of reflux-related hospitalizations in the year before (1.2 ± 1.4) or after (1.4 ± 1.2, P = 0.7) the initiation of transpyloric feeds.
Conclusions: There is more reflux during transpyloric feeding periods than nonfeeding periods, which may explain why some patients continue to have reflux-related hospitalizations even after the initiation of transpyloric feeds.
Transpyloric feeds have been proposed as an alternative to fundoplication in patients with gastroesophageal reflux who are medically complex or in patients with severe or progressive motility disorders. Recent studies have suggested that the clinical benefit to jejunal feeding is comparable with fundoplication benefits and that pulmonary outcomes, particularly aspiration pneumonia, are comparable between jejunal feeding and fundoplication (1,2).
The mechanism for its proposed benefits is that reflux is reduced, possibly as a result of reduced volume of gastric contents; because there is no food in the stomach, there is less volume that can be refluxed. However, this effect may not be related to volume alone, and the exact mechanism of the beneficial effect of transpyloric feeding has not been well studied.
Interestingly, other studies have shown that aspiration pneumonias still occur in patients receiving jejunal feedings, which suggests that transpyloric feeding may not entirely eliminate the gastroesophageal reflux risk. One adult study showed that the rate of gastroesophageal reflux was higher when formula was instilled into the jejunum compared with when saline was instilled, suggesting that jejunal feeding may have an effect on gastroesophageal reflux (3). The rate of reflux reduction in children who are jejunally fed is unknown. It is the goal of the present study to compare the rates of reflux during jejunal feed periods to nonfeed periods.
PATIENTS AND METHODS
We retrospectively reviewed the multichannel intraluminal impedance (pH-MII) tracings of all of the symptomatic patients who had transpyloric feeds in an effort to treat gastroesophageal reflux. All of the patients had pH-MII testing between 2001 and 2010 because of intractable pulmonary or gastrointestinal symptoms. Tube location was confirmed in interventional radiology to be beyond the ligament of Treitz. Each patient's study was divided into feed (when the pump was running) and nonfeed periods (when the feeding pump was not running). We excluded 15 patients from analysis because they were receiving concurrent gastric or oral feeds. Nineteen patients were fed by gastrojejunal (GJ) tubes and 1 was fed by a nasojejunal tube. No patient had a primary jejunal tube.
Reflux, as measured by pH-MII, was classified as acid, nonacid, or pH-only episodes. A reflux episode detected by impedance was defined as a retrograde drop in impedance to more than 50% of baseline in the 2 distal channels. Bolus clearance time was defined as the time from a drop in impedance to 50% of its baseline value to its recovery to 50% of its baseline value in the distal-most impedance channel. Acid reflux episodes are episodes detected by both pH and impedance sensors. Nonacid episodes are episodes detected by impedance sensors only. pH-only episodes are episodes detected by the pH sensor only and are a minimum of 5 seconds in length. Full-column reflux was defined as an episode that reached the highest pair of impedance sensors. The percentage of time that reflux boluses, as detected by pH-MII, are present in the esophagus during feed/nonfeed periods was calculated by dividing the sum of the distal bolus clearance times by the total feed/nonfeed period duration (4).
We calculated the average number of gastroesophageal reflux events per hour during feed and nonfeed periods. We also reviewed the number of reflux-related hospitalizations at Children's Hospital Boston in the year before and the year after the initiation of postpyloric feeds. Hospitalizations were considered reflux related if the discharge summary had a primary or secondary diagnosis of vomiting, reflux, aspiration, aspiration pneumonia, respiratory distress, or pneumonia. All of the patients were maintained on long-term transpyloric feeds in the follow-up period.
Proportions were compared using χ2 analyses, and means were compared using the Wilcoxon ranked sign test. Results are presented as mean ± standard deviation.
The mean age of patients in the present study was 62 ± 59 months; the range was 6 months to 190 months. Indications for pH-MII testing and underlying diagnoses are shown in Table 1. Eighteen patients were taking acid-suppression therapy, 2 were taking baclofen, and 3 were taking erythromycin at the time of pH-MII testing. Sixty percent of the patients had an underlying neurologic diagnosis. Six of 20 (30%) had evidence of aspiration on barium swallow study or nuclear medicine salivagram.
The mean duration of the pH-MII studies was 23.0 ± 2.9 hours. The mean time that feeds were running was 18.0 ± 5.0 hours and the mean time that feeds were discontinued during the 24-hour study was 5.0 ± 4.3 hours. Tables 1 and 2 show the reflux profiles for the 20 patients in the study. There was a significant difference in the total number of reflux episodes during feeds (24.9 ± 25.8) compared with nonfeed periods (3.3 ± 3.2, P = 0.001). The mean number of reflux events per hour was 1.4 ± 1.3 and 0.8 ± 1.1 during feed and nonfeed periods, respectively (P = 0.06). Twenty-nine percent of reflux episodes were full column while the feeds were running, compared with 22% when the feeds were not running (P = 0.6). There was no significant difference in the percentage of time that boluses were present in the esophagus during feed periods (1.0 ± 1.4%) compared with the nonfeed periods (0.6 ± 1.1%, P = 0.5).
In the patients who had regular follow-up at Children's Hospital Boston (n = 17), there was no significant difference in the mean number of reflux-related hospitalizations in the year before (1.2 ± 1.4) or after (1.4 ± 1.2, P = 0.7) the initiation of transpyloric feeds. Eleven of 17 patients (64%) were hospitalized for aspiration pneumonias after transpyloric feeds were initiated; of these 11 patients, 5 had evidence of aspiration on swallow study or salivagram, 3 had no evidence of aspiration on either test, and 3 did not have any radiologic evaluation of swallowing function. To clarify whether there was a characteristic of patients that predicted improved hospitalizations after transpyloric feeding, we compared the reflux profiles of patients who did and did not have a reduction in hospitalizations after the initiation of transpyloric feeds. There were no differences in the reflux profiles in patients who did and did not have a reduction in hospitalizations (Table 3). Additionally, there was no difference in the mean symptom indices in patients who did and did not have a reduction in hospitalizations after the initiation of transpyloric feeds.
We then determined the difference in the change in hospitalizations between patients with and without neurologic compromise. Patients who had neurologic compromise had an average increase in hospitalizations in the year after transpyloric feeds were initiated (0.8 ± 1.5 hospitalizations) compared with a reduction in average hospitalizations after transpyloric feeds were initiated in patients without neurologic compromise (−0.3 ± 1.0 hospitalizations), but this difference was not statistically significant (P = 0.1). There were no differences in the reflux profiles between patients with and without neurologic compromise (P > 0.6).
Two (ages 26 and 32 months) of the 20 patients underwent fundoplication after the initiation of transpyloric feeds; the total number of reflux events per 24-hour period in these patients was 11 and 28. Both improved symptomatically after the fundoplication, despite a normal reflux burden during transpyloric feeds; 1 patient, who was neurologically compromised, had an abnormal swallow study with aspiration, and 1, who was neurologically normal, had no evidence of aspiration on swallow study.
There were more total reflux events and a higher number of reflux events per hour during transpyloric feeding compared with nonfeeding periods. This is consistent with adult studies that have shown, using pH analyses, that infusion of formula into the jejunum results in increased esophageal acid exposure and an increased number of reflux events (3). Studies in adults have suggested that the mechanism for this increased acid exposure is caused by increases in the number of transient lower esophageal sphincter relaxations in patients with evidence of reflux disease; instillation of fat into the small bowel increased the transient lower esophageal sphincter relaxations, which resulted in increased acid reflux as measured by pH probe (5). The continued presence of reflux during transpyloric feeding may explain why patients continue to have reflux-related hospitalizations after transpyloric feeds are initiated and why there appears to be no difference in the present study between the mean number of reflux-related hospitalizations before and after initiation of transpyloric feeds.
Although reflux is clearly still occurring during transpyloric feeding, the mean number of total reflux events detected by pH-MII is still significantly less than the total reflux events per 24-hour period that we have previously reported in patients with significant gastroesophageal reflux disease (GERD) symptoms in whom the mean number of reflux events per 24 hours was 54.9 ± 17.9 (P = 0.003) (6). Interestingly, the percentage of full-column reflux events in patients with transpyloric feeding is significantly lower than we have previously reported in GERD patients who had a mean of 30% ± 15% of full-column events (vs 18% ± 16% in the present study, P = 0.04) (6). This suggests that, although there is an increased rate of reflux while the pump is running, the overall number of reflux events and the percentage of full-column events is lower in patients with transpyloric feeding than in patients with GERD, which may explain why some patients experience symptomatic improvement with transpyloric feeds (6). Although none of the patients in the present study are neonates, this reduction in full-column events may explain the findings in 2 neonatal studies that have shown a reduction in apnea and bradycardia events in preterm infants who were transpylorically fed (2,7). Interestingly, pH-MII studies have failed to show a correlation between apnea and bradycardia and reflux events, which suggests that either pH-MII is a flawed method of correlating reflux with events or that the mechanism of apnea and bradycardia cannot entirely be explained by full-column reflux into the esophagus but by another possible mechanism such as gastric distension (8,9).
To address the issue of whether pH-MII results predicted clinical outcome, we looked at the reflux profiles of patients who did and did not experience a reduction in hospitalizations after initiation of transpyloric feeding to determine whether patients who had a reduction in hospitalizations had a different reflux profile than those who did not have a reduction; there were no differences in the reflux profiles, the symptom indices, and the proportion of patients with evidence of aspiration by radiologic imaging between these 2 groups. The persistently high rates of hospitalizations after initiation of transpyloric feeding could be a result of 2 different factors: the patient's disease was independent of reflux or any remaining reflux, although not pathologic based on the number of events or symptom association, was responsible for their disease.
To clarify whether the patient's symptoms or disease is reflux related, ideally we would have performed pH-MII studies in all of the patients, not just persistently symptomatic ones, before and after the initiation of transpyloric feeds to see whether the reflux burden changed and the symptoms improved. Unfortunately, patients rarely consent to 2 pH-MII studies and, thus, as in the present study, we only performed pH-MII testing on patients who remained symptomatic despite transpyloric feedings, biasing the results toward diseases or symptoms that are unrelated to reflux.
The second factor, that any remaining reflux present after the initiation of transpyloric feeds can cause disease, is addressed in 2 ways. First, we would expect that patients with evidence of aspiration on swallow study or salivagram would be at risk for reflux-related hospitalizations after transpyloric feeding because any full-column reflux into the oropharynx could be aspirated owing to impaired airway protective mechanisms. We found that 33% of patients who experienced a reduction in hospitalizations had evidence of aspiration on swallow study compared with 67% of patients who did not experience a reduction. Although these proportions were not statistically significant because of the small sample size, we believe that aspiration either of oropharyngeal secretions or of residual reflux may be a risk factor for a worse prognosis with transpyloric feeds and that thought should be given to altering the jejunal feeding regimens of patients who aspirate. Nocturnal jejunal feeds may need to be reduced or avoided to prevent recumbent aspiration events in the patient who remains symptomatic with transpyloric feeds. Second, 2 patients who continued to have pulmonary symptoms despite transpyloric feedings underwent fundoplication and both improved symptomatically after the fundoplication, suggesting that the residual reflux with transpyloric feeding may still be problematic, although we acknowledge that many fundoplications may have been prevented in the other patients because of the transpyloric feeds.
Our high rate of reflux-related hospitalizations after transpyloric feeds has been echoed by other investigators. Srivastava et al found that 16% of patients with gastrojejunal tubes developed aspiration events (1). Wales et al found that 31% of their patients developed aspiration pneumonias after gastrojejunal tube placement (10). Clinically, we use impedance and transpyloric feed trials as 2 diagnostic tools to determine reflux-related lung disease. However, as our study has shown, neither pH-MII nor a transpyloric feeding trial can perfectly predict which patients have reflux-related lung disease, and fundoplication may be therapeutic even with a normal reflux burden as determined by pH probe and MII normal values and with a failed trial of transpyloric feeds (6,11,12). As a result, new, more specific tools are needed to determine reflux-related disease.
In summary, there is still evidence of reflux during transpyloric feedings and the rate of reflux during feeding times is double that of nonfeeding periods. Although transpyloric feeding is helpful in controlling reflux in complex patients, gastroesophageal reflux still occurs despite this method of feeding and aspiration events and reflux-related hospitalizations are still possible despite transpyloric feeds.
1. Srivastava R, Downey EC, O'Gorman M, et al. Impact of fundoplication versus gastrojejunal feeding tubes on mortality and in preventing aspiration pneumonia in young children with neurologic impairment who have gastroesophageal reflux disease. Pediatrics 2009; 123:338–345.
2. Misra S, Macwan K, Albert V. Transpyloric feeding in gastroesophageal-reflux-associated apnea in premature infants. Acta Paediatr 2007; 96:1426–1429.
3. Lien HC, Chang CS, Yeh HZ, et al. The effect of jejunal meal feeding on gastroesophageal reflux. Scand J Gastroenterol 2001; 36:343–346.
4. Srinivasan R, Vela MF, Katz PO, et al. Esophageal function testing using multichannel intraluminal impedance. Am J Physiol Gastrointest Liver Physiol 2001; 280:G457–G462.
5. Holloway RH, Lyrenas E, Ireland A, et al. Effect of intraduodenal fat on lower oesophageal sphincter function and gastro-oesophageal reflux. Gut 1997; 40:449–453.
6. Rosen R, Furuta G, Fritz J, et al. Role of acid and nonacid reflux in children with eosinophilic esophagitis compared with patients with gastroesophageal reflux and control patients. J Pediatr Gastroenterol Nutr 2008; 46:520–523.
7. Malcolm WF, Smith PB, Mears S, et al. Transpyloric tube feeding in very low birthweight infants with suspected gastroesophageal reflux: impact on apnea and bradycardia. J Perinatol 2009; 29:372–375.
8. Peter CS, Sprodowski N, Bohnhorst B, et al. Gastroesophageal reflux and apnea of prematurity: no temporal relationship. Pediatrics 2002; 109:8–11.
9. Mousa H, Woodley FW, Metheney M, et al. Testing the association between gastroesophageal reflux and apnea in infants. J Pediatr Gastroenterol Nutr 2005; 41:169–177.
10. Wales PW, Diamond IR, Dutta S, et al. Fundoplication and gastrostomy versus image-guided gastrojejunal tube for enteral feeding in neurologically impaired children with gastroesophageal reflux. J Pediatr Surg 2002; 37:407–412.
11. Shay S, Tutuian R, Sifrim D, et al. Twenty-four hour ambulatory simultaneous impedance and pH monitoring: a multicenter report of normal values from 60 healthy volunteers. Am J Gastroenterol 2004; 99:1037–1043.
12. Rudolph CD, Mazur LJ, Liptak GS, et al. Guidelines for evaluation and treatment of gastroesophageal reflux in infants and children: recommendations of the North American Society for Pediatric Gastroenterology and Nutrition. J Pediatr Gastroenterol Nutr 2001; 32(suppl 2):S1–S31.