Gastroesophageal Reflux During Gastrostomy Feeding

Transcription

1 GSTRONTROLOGY 1994;106:13-18 Gastroesophageal Reflux During Gastrostomy Feeding ROBRT M. COBN*, LN WINTRUB,* NTHONY J. DIMRINO, JR.,* and SIDNY COHN* *Department of Medicine and Division of GastroenterologY, Presbyterian Medical Center, University of Pennsylvania, Philadelphia; and *Department of Medicine, and Division of GastroenterologY, Temple University, Philadelphia, Pennsylvania Background~ims: spiration pneumonia is one of the most serious complications of gastrostomy tube feeding, with a reported incidence of 10%-20% in nursing home patients. The aims of this prospective study were to examine lower esophageal sphincter (LS) pressure before and after placement of gastrostomy tubes and to examine the effects of rapid intragastric bolus and slow, continuous feeding on LS pressure. Methods: Ten subjects were enrolled in the study. Basal LS pressure was measured before and after placement of gastrostomy tubes. Thereafter, LS pressure was measured for 15 minutes during rapid intragastric infusion of 250 ml of aft enteral feeding formula and 100 ml water and continuous infusion of the enteral feeding formula at 80 ml/h. Scintigrams evaluating gastroesophageal reflux were obtained during each method of feeding. Results: Placement of gastrostomy tubes had no effect on basal LS pressure. Rapid intragastric bolus infusion led to a reduction in LS pressure to incompetent levels at 2.1 _+ 2.0 mm Hg (P < 0.001). Free gastroesophageal reflux to the sternal notch was shown by scintigraphy. Slow, continuous gastrostomy feedings did not alter LS pressure (P > 0.05) or show free gastroesophageal reflux by scintigraphy. Conclusion~: Gastroesophageal reflux and aspiration in patients fed via the gastrostomy tube may be caused by LS relaxation secondary to gastric distention caused by distention of the stomach. p ercutaneous endoscopic gastrostomy tube feeding has become the most popular method of enteral feeding for debilitated patients. ~ In 1990, it was estimated that there were approximately 800,000 patients in the United States who have been fed via enteral feeding tubes. 2 lthough effective, there are a number of reported complications. Potential complications include wound infections, stoma/leaks, pneumoperitoneum, tube dislodgement, perforation or peritonitis, bleeding, and gastrogolic fistula. 3 The most common serious complication is that of aspiration pneumonia with a reported incidence of 10%-22%. 4 This complication is more frequent in patients in nursing homes who are fed by gastrostomy tubes. There are several factors that could influence the risk of aspiration (with consequent pneumonia and possible respiratory failure) in gastrostomy-fed patients. These factors include impaired esophageal clearance mechanisms (peristalsis), decrease in lower esophageal sphincter (LS) pressure with associated incompetence, volume or rate of gastric infusion through the gastrostomy tube, rate of gastric emptying, posture during feeding, associated diseases such as diabetes or scleroderma, use of drugs that may further alter LS tone, and the effect of prior surgery (such as vagotomy) on LS function. Reflux of gastric contents leading to aspiration is in part dependent on the effectiveness of the LS to act as a barrier to gastroesophageal reflux. Whereas several studies have investigated the response of LS to intragastric balloon distention, ~ infusion of intragastric amino acids, 6 effects of gastrostomy on gastroesophageal reflux, 7 and gastric emptying, s the competence and response of the LS pressure to direct bolus and continuous infusion of liquid food into the stomach has not been studied. The purpose of this prospective study was to examine the effect on LS pressure ofgastrostomy tube placement and to quantify the effects of bolus and continuous feeding through the gastrostomy tube on LS pressure and gastroesophageal reflux. Materials and Methods Patients Ten patients referred for endoscopic placement of a gastrostomy tube were studied. The group included five men and five women (mean age, Z-_ 9.0 yrs; range, yrs). Diagnoses of patients included eight with cerebrovascular accident and two with dementia. Institutional Review Board approval was obtained on December 12, 1989, before initiating the study. Informed consent was obtained from all patients and/or family members in all cases. Study Design Before placement ofgastrostomy tubes, esophageal ma- t nometry was performed to determine basal LS pressure. ll 1994 by the medcan Gastmenterologlcal ssociation /94/$3.00

2 14 COBN T L. GSTRONTROLOGY Vol. 106, No. 1 patients underwent oral passage of a multilumen dent sleeve/ side hole manometric assembly 9 water perfusion esophageal motility catheter. The manometric orifices in the 8 lumen catheter were 5 cm apart at a 90 angle (diameter, 4.8 mm; internal diameter of each, 0.8 mm) with a sleeve length of 5.0 cm. pneumohydraulic capillary infusion system (rndorfer Medical Specialties Inc., Greendale, WI) was used for continuous infusion of each lumen at a rate of 0.5 ml/min. ach manometric catheter was connected to a transducer and in turn connected to a direct writing recorder (Dynograph Recorder R611, sophageal Motility System; Sensor Medics, naheim, C). The compliance of the system was low (400 mm Hg/s). 7 LS was localized via the station pull-through technique. LS pressure for each patient was measured from the mean gastric pressure to the end expiratory pressure. ll patients were fasted and off of all medications known to alter LS pressure 12 hours before the study. Patients were kept in a supine position with the head pf the bed elevated approximately 30 during the course of manometry. Following esophageal manometry, all withheld medications were restarted after gastrostomy placement. ~astrostomy placement was performed using the Bard (Tewksbury, MD) percutaneous endoscopic gastrostomy tube assembly and standard technique, l ll patients had normal endoscopic findings with the site of gastrostomy placement located approximately one third the distance from the left costal margin at the midclavicular line to the umbilicus. There were no complications from the procedure. Repeat ~evaluation of LS pressure was performed 24 hours after gastrostomy placement. Patients were fasted and off of all medications known to affect LS pressure 12 hours before esophageal manometry. LS was measured with patients lying supifie and the head of the bed elevated approximately 30 during the course of manometry. First, LS pressure was measured continuously for 15 minutes following a rapid bolus infusion of 1 can (250 ml) of Jevity (Ross, Columbus, OH) (1.06 cal/ml; osmolality, 310 mosm/kg; protein, 16.7%; carbohydrate 53.3%; fat, 30%) followed by 100 ml water. Two hours later, LS pressure was measured continuously for 15 minutes during a continuous infusion of Jevity at 80 ml/h. Five patients were evaluated using gastroesophageal reflux scintigraphy. Scintigrams were obtained with patients lying supine and the head of the bed at a 30 angle. First, 1 can of Jevity (250 ml) was labeled with 5.2 mci technetium-99 sulfur colloid. rapid bolus infusion of this was given via the gastrostomy tube followed by 100 ml water. $cintigrams were obtained to determine gastroesophageal reflux every 10 seconds over a 15-minure period. Twenty-four hours later, 1 can of Jevity (250 ml) was labeled with 5.2 mci of rechnerium Tc 99m sulfur colloid and infused continuously via the gastmstomy tube at 80 ml/h. Scintigrams were obtained every 10 seconds over a 15- minute period to show gastroesophageal reflux. Statistical nalysis LS pressure was compared l~efore and after gastrostomy placement, before and after bolus feeding, and before O} '-I- V CO w 10-._1 5 + PR +!! m POST Rgure 1. Mean pregastrostomy (15.9 _+ 4.0 mm Hg) vs. mean postgastrostomy basal LS pressure (16.6 _+ 4.2 mm Hg). P < 0.1; n : 10. and after continuous feeding. ll pressures are expressed as mean + SM. The paired Student's t test was used to determine statistical significance. Results In Figure 1, mean pregastrostomy LS pressure was mm Hg compared with a mean postgastrostomy LS pressure of mm Hg. There was no significant difference between these pressures (P > 0.05). Figure 2 compares LS pressure before and after rapid intragastric bolus gastrostomy feeding using 250 ml Jevity followed by 100 ml water. Mean prebolus LS pressure ( mm Hg) was markedly reduced to mm Hg (P < 0.001). Figure 2B shows a characteristic manometric tracing of LS pressure men-

3 January 1994 BOLUS GSTROSTOMY FDING RDUCS LS PRSSUR \ 20 I 15 uj -J 10 \ 5 BSL BOLUS B i~ 'I" O g mln. 18 mln. ~BOLUS~ TIM (2.Smmlsec) Rgure 2. () Mean basal (16.6 _ 4.2 mm Hg) vs. mean bolus gastrostomy LS pressure ( mm Hg). P < 0.001, n = 10 for an average of _ 30.1 seconds. (B) Bolus feeding manometric tracing for 15 minutes. sured for 15 minutes following a rapid intragastric infusion of Jevity followed by water. The reduction in LS pressure was reduced for (mean + SM) seconds in all 10 subjects. In Figure 3, the LS pressure before angl after Jevity at 80 ml/h is compared. Mean LS pressure before infusion was mm Hg compared with mm Hg during infusion. This change was not significant (P ~ 0.05). Figure 3B shows a manometric tracing during continuous feeding. Scintigraphy Figure 4 shows a scintigram of a patient during bolus feeding. The completion of the bolus feeding is at time zero. t I minute and 30 seconds postbolus feeding, there was gastroesophageal reflux to the sternal notch. t 2 minutes and 30 seconds after bolus feeding, gastroesophageal reflux persists but to a lesser degree. t 15 minutes, no reflux is shown. Pulmonary aspiration did not occur in any patient.

4 16 COBN T L. GSTRONTROLOGY Vol. 106, No o'j -r 15 O0 LU. 10 \ 5.'~ 3o ~10 iii -I I BSL CONTINUOUS B CONTINUOUS TIM (2.5mm/sec) In Rgure 3. () Mean basal (16.6 _+ 4.2 mm Hg) vs. mean continuous gastrostomy feeding LS pressure (14.3 _+ 4.3 mm Hg). P < 0.05; n : 10. (B) Continuous feeding manometric tracing for 15 minutes. Figure 4B shows scintigrams during continuous feeding via the gastrostomy tube. For a period of 15 minutes, no gastroesophageal reflux is shown. Discussion Gastrostomy tubes have replaced nasogastric tubes as the most common form of enteral feeding in debilitated patients. The incidence of aspiration pneumonia in nursing home patients fed via the gastrostomy tube has been reported to be 22.9%. In a large series, seven deaths (6% Of the patients) were attributed to aspiration pneumonia. 4 The risk of aspiration pneumonia in patients :fed via gastrostomy tubes in a neurosurgical intensive care unit was compared during gravity and slow, continuous enteral feedings. I~ There was no significant difference in the incidence of aspiration pneumonia. The discrepancy in these observations is not apparent. The methodology of feeding may be an important'factor in contributing to the risk of aspiration pneumonia. major determinant of gastroesophageal reflux is the competence of the LS. Our results indicate that LS pressure is slightly increased after gastrostomy placement. It has been suggested that this small increase in LS pressure might be caused by a "gastropexy effect" after the gastrostomy is placed in the anterior apposition to the gastric wall. 7

5 January 1994 BOLUS GSTROSTOMY FDING RDUCS LS PRSSUR 1 0 min B < s t e r n a l notch rnln 20 < s t e r n a l notch sec I min 20 0 mln sec t, ~.- min 30 s e c i~:" 17 m~ r 15 mln 2 mln 38 sec 15 mln Figure 4. () Scintigraphy depicting free gastroesophageal reflux during bolus gastrostomy feeding (250 ml Jevity plus 100 ml water labeled with 5.2 mci of 99m'rc-sulfur colloid) for 15 minutes. (B) No gastroesophageal reflux was seen during continuous gastrostomy feeding, (80 ml/h Jevity). The major objective of this study was to determine the effect of the rate of infusion of gastric feedings on LS pressure and the induction of gastroesophageal reflux. Slow, continuous feeding had no effect on LS pressure, whereas rapid intragastric bolus feeding reduced LS pressur~ to incompetent levels. Free gastroesophageal reflux, as shown by scintigraphy, corresponded to reduced LS pressure. Inhibition in LS pressure may occur by a reflex response during gastric fundal distention. Gastric fundal distention in cats and dogs has been shown to decrease LS pressurej 2''3 linear decrease in LS pressure has been reported with increased fundic pressures in humansj 3 marked reduction in LS pressure to incompetent levels has been noted after distention 14 of the stomach with CO2 gas. The association of gastroesophageal reflux with gastric distention has been attributed to an increase in transient lower esophageal sphincteric relaxation. 5 This phenomenon is probably a contributing factor to gastroesophageal reflux, but is not evaluated in this study., In conclusion, our study indicates that rapid intragastric bolus feeding via the gastrostomy tube causes transient relaxation of the LS associated with gastroesophageal reflux to the sternal notch. This finding was not shown during feeding by continuous infusion via the gastrostomy tube. This study suggests that the practice by which patients are fed via gastrostomy tubes be revised so that the risk of aspiration pneumonia can be reduced. References 1. Gauderer MW, Ponsky JL, Izant RJ. Gastrostomy without laparoscopy: a percutaneous endoscopic approach. J Pediatr Surg 1980; 15: Mobarhan S, Trumbore LS. nteral tube feeding: a clinical perspective on recent advances. Nutr Rev 1991;49: Mamel JJ. Percutaneous endoscopic gastrostomy. m J Gastroenterol 1989;84: Cogen R, Weinryb J. spiration pneumonia in nursing home patients fed via gastrostomy tubes. m J Gastroenterol 1989; 84: Holloway RH, Hongo M, Berger R, McCallum RW. Gastric distention: a mechanism for postprandial gastroesophageal reflux. Gastroenterology :1.985;89: McCallum RW, Kuljian B, Holloway RH, Walsh JH. ffect of intragastric amino acids on lower esophageal sphincter pressure and serum gastrin in man. m J Gastroenterol 1986;81: Johnson D, Hacker JF, Benjamin JB, Casteglio, Chobanian JJ, Vanness MM, Catlaw L. Percutaneous endoscopic gastrostomy effects on gastroesophageal reflux and the lower esophageal sphincter. m J Gastroenterol :1.987;82: Kutcher W, Cohen LB, hrlich L. Changes in gastric emptying of liquids following percutaneous endoscopic gastrostomy (PG) (abstr). m J Gastroenterol 1989;84: Dent J. new technique for continuous sphincteric pressure measurement. GastroenteroloSy 1976;71: Ponsky JL, Gauderer MW, Stellato T, szodi. Percutaneous

6 18 COBN T Lo GSTRONTROLOGY VoL 106, No. 1 approaches to enteral alimentation. m J Surg 1985; 149: Kocan M J, Hickisch SM. comparison of continuous and intermittent enteral nutrition in NICU patients. J Neurosci Nuts 1986; 18: Jennewein HM, Hummelt H, Siewert R, Weiser F, Walcheck F. The effect of pressure changes inside the antrum on lower esophageal sphincter (LS) pressure in man and dog. Hepatogastroenterology 1976; 23: Muller-Lissner J, Blum. Fundic pressure rise lowers lower esophageal sphincter pressure in man. Hepatogastroenterology 1982; 29: Wyman JB, Dent J, Heddle R, Dodds WJ, Toouli J, Downton J. Control of belching by the lower esophageal sphincter. Gut 1990;31: Received September 9, ccepted ugust 31, ddress requests for reprints to: nthony J. DIMarlno, Jr., M.D., Presbyterian Medical Center, Suite 218, Wflght Building, 39th and Market Streets, Philadelphia, Pennsylvania