Mechanism of gastroesophageal reflux in premature infants with chronic lung disease

Mechanism

of Gastroesophageal Reflux in Premature Chronic Lung Disease

By T. Omari,

C. Barnett,A.

Snel, G. Davidson, Adelaide,

Background/furpose:The ize the motor events (GER) and esophageal infants with chronic

aim of this study was to characterresponsible for gastroesophageal reflux acid clearance in a cohort of preterm lung disease (CLD).

Methods: Esophageal motility was recorded for 2 to 3 hours postprandially in 14 preterm infants with mild-moderate or moderate-severe CLD at 33 to 39 weeks postmenstrual age using a sleeve-side hole micromanometric assembly into which a pH probe had been installed. Results: Twenty-six acid GER episodes were recorded by pH probe. Of these, 22 (85%) were associated with transient lower esophageal sphincter (LES) relaxation. Transient LES relaxations were significantly longer in duration than single swallows (13.7 seconds v 4.6 seconds, P< .OOl) and had lower nadir pressures (1.1 mm Hg v 2.5 mm Hg, P< ,001).

ASTROESOPHAGEAL REFLUX (GER) disease is present in 18% to 23% of infants suffering from chronic lung disease (CLD) or bronchopulmonary dysplasia.‘J Although not the primary cause of CLD, GER may exacerbate CLD because of an increased risk of aspiration, a pathophysiological association clearly supported by the therapeutic benefit of fundoplication on the respiratory status of infants with concomitant GER and CLD.1-7 In infants with CLD, increased inspiratory effort, coupled with transient increases in intraabdominal pressure related to coughing and crying, may increase the likelihood of the occurrence of GER.g In addition, commonly used respiratory stimulants such as theophylline can stimulate gastric acid secretion,9 decrease lower esophageal sphincter (LES) pressurelOJ and increase the incidence of acid reflux and reflux symptoms.ll Other therapeutic measures including frequent chest physiotherapy and supine positioning may also exacerbate reflux.‘2-14 Esophageal motor function and the mechanisms of GER have been relatively unstudied in infants and children with CLD. Manometric side hole “pullthrough” estimates of LES pressure in children with CLD range from 11 to 12 mm Hg.4,15Although this is lower than in normal subjects (usually 15 to 30 mm Hg), these pressures are sufficient to maintain esophagogastric competence. In adults and children with or without GER disease, transient LES relaxation is the predominant mechanism of acid GER.16-22Through the development

G

Journal

ofPediatricSurgery,Vol

34, No 12 (December),

1999: pp 1795-1798

R. Haslam,

M. Bakewell,

Infants

With

and J. Dent

Australia

During periods of peristaltic esophageal acid refluxate more wave sequences.

esophageal acidification (pH body pressure wave sequences effectively than nonperistaltic

below 4), cleared pressure

Conclusions: In preterm infants with CLD, transient LES relaxations are the predominant mechanism underlying GER, and esophageal clearance mechanisms are fully functional, which is similar to that seen in healthy preterm infants. J Pediatr Surg 34:1795-1798. Copyright o 1999 by W.B. Saunders Company. INDEX WORDS: gastroesophageal acid clearance, ation.

Chronic reflux, transient

lung disease, preterm neonate, acid esophageal motility, esophageal lower esophageal sphincter relax-

of novel feeding catheters, which include multiple miniature manometric channels, we recently have determined that transient LES relaxations also occur in healthy premature infants as young as 32 weeks’ postmenstrual age.‘3 As with older subjects, transient LES relaxations also were the predominant reflux mechanism in healthy preterm infants.23.24To date, there are no similar data in premature infants suffering from CLD. The aim of this study was to use combined manometry and pH recording to characterize the motor mechanisms responsible for GER and esophageal acid clearance in these infants. MATERIALS

AND METHODS

Sztbjects Infants with CLD were defined as those who had sustained need for oxygen at or beyond 28 postnatal days and who had characteristic features of CLD on chest radiograph.25 Fourteen gavage-fed pretem infants (9 boys. 5 girls) with CLD were studied. The severity of lung

From the Gastroenterology Unit, Neorzatal Medicine Unit, and Departmenr of Biomedical Engineering, Women S and Children S Hospital, North Adelaide, Australia, and Gastrointestinal Medicine, Royal Adelaide Hospital, Adelaide, Australia. Supported by grants from the National Health and Medical Research Courzcil of Australia, WornenS and Childrenk Hospital Research Foundation, and Channel 7 Children’s Uesearch Foundation of South Australia. Address reprint requests to Dr TI. Omari, Gasrroenterology Unit, Women k and Children k Hospital, North Adelaide, Australia, 5006. Copyright 0 1999 by WB. Sauders Company 0022.3468/99/3412-0010$03.00/0 1795

1796

disease at the time of manometric study was determined using an established scoring system for assessment of bronchopulmonary dysplasia.s6 Of the 14 infants, 7 were assessed as having moderate-severe CLD, whereas the remainder were assessed as having mild-moderate CLD. The mean postmenstrual age of the infants was 37 weeks (range, 33 to 39 weeks) with a mean postnatal age of 75 days (range, 50 to 104 days). Infants had no evidence of neurological dysfunction and were not on prokinetic medication. Mean infant weight was 2,099 g (range, 1,290 to 2,720 g). Five infants were receiving either theophylline or caffeine treatment. Twelve infants were being gavage (tube) fed with expressed breast milk and 2 with infant formula (Enfalac 20 calorie, Mead Johnson, Canada). Feedings had a pH of 7.0 to 7.1 and feeding volume ranged from 18 to 80 mL (mean, 50 mL). The study protocol was approved by the Ethics Research Committee of the Women’s and Children’s Hospital and written informed parental consent was obtained before each study.

Manornetric Technique Patterns of esophageal motility were recorded using a micromanometric feeding assembly (2 mm outside diameter), which incorporated a 2..5-cm sIeeve sensor for LES pressure measurement?‘~24,*7 Six sidehole sensors (located at -1.5 cm, 1.5 cm, 3.0 cm, 4.5 cm, 9.5 cm, and 10.5 cm relative to the sleeve midpoint) recorded pressures from the pharynx, esophageal body, and stomach. Pharyngeal side holes were air perfused at 5 mL/min, and all other side holes were perfused with sterile degassed water at 0.02 mL/min (esophageal) and or 0.04 ml/mm (sleeve and stomach) using hydraulic resistors (Dentsleeve Pty, Wayville, South Australia). Bubble entrapment was minimized by preflushing of the catheter lumina with carbon dioxide.28 Esophageal pH was monitored with an antimony pH probe (outside diameter, 1.5 mm; Synectics, Stockholm, Sweden) positioned 3 cm proximal to the center of the LES sleeve. Analogue pressure transducer and pH probe signals were amplified and filtered with a BME 1600 polygraph (Department of Biomedical Engineering, Women’s & Children’s Hospital, North Adelaide, South Australia). Data acquisition and analysis were performed on a Macintosh Quadra 700 with software based on National Instruments’ Labview (M.A.D. software, Royal Adelaide Hospital, C. Malbert).

Protocol The assembly was passed transnasally and positioned with the midpoint of the sleeve straddling the LES high-pressure zone. After positioning and with the infant in the right lateral position, the feeding was administered over 15 to 30 minutes via the core channel of the assembly. Esophageal pH and spontaneous esophageal body and LES motor patterns were then recorded during the feed and then for 2 to 3 hours postprandially (mean, 2.6 hours). The assembly was well tolerated by the infants with no adverse effects occurring related to the procedure.

Analysis of Manometric Tracings The postprandial manometric recordings were analyzed to assess esophageal and LES motor function as well as mechanisms of GER and GER clearance. IdentijTcation of GER episodes and evaluation of esophageal clear-ante mechanisms. Acid GER was defined as a drop of esophageal pH from above to below 4 for at least 5 seconds or a drop of more than 1 pH unit, if pH was already below 4. The esophageal and LES motor patterns associated with the onset and resolution of each acid reflux episode was characterized. The acid clearance time was defined as the time taken for esophageal pH to return to a level of 4 or more. The patterning of esophageal body pressure waves that exceeded a threshold amplitude of 10 mm Hg was defined as peristaltic or nonperistaltic (synchronous,

OMAN

ET AL

retrograde, or incomplete) according to criteria already established for premature infants.‘3~21.27 Pressure waves were classified as swallow induced if a swallow was recorded in the 5 seconds before the onset of the most proximal contraction, or swallow unrelated if unassociated with swallowing. Pressure wave sequences that occurred within 15 seconds of the previous swallow were not analyzed because of the inhibitory effects of prior peristalsis. Determination of LESpressure and characterisation of LES relaxufiorz. LES pressure was defined as the difference between endexpiratory LES pressure and gastric pressure. Single swallow-related LES relaxations were defined as those within 2 seconds before and 4 seconds after the onset of the upstroke of the pharyngeal pressure wave.Z3z24 Multiple swallowing was classified as a sequence of 2 or more swallows that started within 2 seconds of the onset of the relaxation with an interswallow interval of less than 5 seconds. LES relaxations that occurred independently of swallowing (>2 seconds before or >4 seconds after pharyngeal contraction) with relaxation rates of ~1 mm Hg/s were classified as transient LES relaxationsz2-s1 Each LES relaxation also was analyzed for nadir pressure and duration of relaxation according to established criteria.‘3,24 Straining was identified by sustained increases (25 mm Hg) in intragastric pressure for at least 10 seconds associated with a corresponding rise in esophageal body pressure. The computerized data analysis system made it possible to recognize the occurrence of LES relaxation during straining by comparison of the pressure differential between the gastric sensor and the sleeve.

Statistical Analysis Grouped data were compared using an analysis of variance (ANOVA) technique (F-test and Scheffe’s test). Paired data were compared with a paired t test, and correlation was determined by simple regression analysis. A P value of less than .05 was considered statistically significant.

RESULTS

Mechanisms of Acid GER and Esophageal Volume Clearance Twenty-six acid GER episodes were recorded in the infants studied. Ninety-six percent of these reflux episodes were associated with LES relaxation. These were mostly transient LES relaxations, but occasionally LES relaxations triggered by swallowing with or without peristalsis were associated with reflux (Table 1). Eleven acid reflux episodes occurred during periods of abdominal straining. Of these, 8 were triggered by transient LES relaxation, 3 during swallow-induced relaxation, and 1

Table 1. Esophageal Acid Gastroesophageal

Transient

Motor Events Recorded in Association With Reflux Episodes in Premature Infants With Chronic Lung Disease

GER Mechanism

No. of Acid GER Episodes 1%)

LES relaxation

22 (84.6)

Swallow-induced LES relaxation Single swallow Multiple swallow Peristaltic failure Hypotonic LES Total

2 1 1 26

(7.7) (3.8) (3.81 (100)

REFLUX

IN INFANTS

WITH

CHRONIC

LUNG

1797

DISEASE

occurred when resting LES pressure was less than 2 mm Hg indicating LES hypotonicity. Overall, the mean rate of acid GER was 0.8 t 0.3 episodes per hour. The effect of feeding type could not be evaluated because all but 2 infants received expressed breast milk. Acid GER occurred at a rate of 1.0 2 0.6 episodes per hour in the 5 infants who received methylxanthines compared with 0.6 + 0.4 episodes per hour in the infants who did not receive this therapy; however, this difference did not reach statistical significance (P = ,115). Acid GER rate was not altered by the severity of disease (0.7 +- 0.5 and 0.8 t 0.4 episodes per hour for mildmoderate and moderate-severe disease, respectively). The rate of acid GER did not correlate with postmenstrual age (r = 0.13, not significant). For those acid reflux episodes identified by a pH drop from above to below 4, the mean clearance time of these acid reflux episodes was 118 t 96 seconds (range, 15 to 1,161 seconds). The esophageal pressure wave sequences recorded during these periods of esophageal acidification were usually swallow-induced peristaltic (5 1%) or swallow-unrelated nonperistaltic (predominantly synchronous and incomplete; 37%). Pressure waves peristaltic in sequence (both swallow induced and swallow unrelated) were most effective in facilitating acid clearance, resulting in a mean increase in esophageal pH of 0.6 2 0.1 units per sequence. In contrast, nonperistaltic types of pressure wave sequence produced significantly smaller increases in pH of 0.28 + 0.1 (P < .Ol) and 0.1 2 0.0 (P < .OOl) per sequence for synchronous and incomplete types, respectively. Resting LES Pressure and Characteristics of LES Relaxation The mean resting LES pressure for all infants was 14.4 t 2.5 mm Hg. LES pressure was not altered by methylxanthine therapy (16.3 2 3.2 mm Hg in infants not receiving therapy v 14.4 ? 2.5 mm Hg receiving therapy, not significant), and infants with mild-moderate disease had similar LES pressures to those with moderatesevere disease (14.8 2 3.8 mm Hg v 14.1 -t 3.5 mm Hg, respectively). Resting LES pressure did not correlate with postmenstrual age (r = 0.45, not significant). Transient LES relaxations and multiple swallowinduced relaxations were longer in duration than single swallow-induced LES relaxations (13.7 t 5.0 seconds and 9.9 i 3.9 seconds v 4.6 ? 0.7 seconds, respectively; P < ,001). Nadir pressures also were lower for transient LES relaxation and multiple swallow-induced relaxation compared with single swallow-induced LES relaxation (1.8 + 1.5mmHgandl.l 2 1.6mmHgv2.5 2 1.7mm Hg, respectively; P < .Ol). Overall, transient LES relaxations occurred at an average of 2.3 t 0.5/h (range, 0.0 to 7.0/h). The mean rate of transient LES relaxations was

3.2 ? 0.8/h in the first study hour and 2.0 t- 0.4/h during the remainder of the study (P = .09). Transient LES relaxations occurred at a rate of 3.1 2 1.1 episodes per hour in the 5 infants who received methylxanthines compared with 1.8 2 0.3 episodes per hour in the infants who did not receive this therapy; however, this difference did not reach statistical significance (P = .22). The occurrence of transient LES relaxations was not altered by the severity of disease (1.8 ? 0.5 and 2.8 t 0.7 episodes per hour for mild-moderate and moderatesevere disease, respectively; P = .33). The rate of transient LES relaxations did not correlate with postmenstrual age (r = 0.22, not significant). DISCUSSION

The results of this study show that in preterm infants with CLD, transient LES relaxations are the most common LES motor pattern associated with postprandial acid reflux episodes. In these babies, the LES is a competent antireflux barrier, which relaxes appropriately with hypopharyngeal swallow. Esophageal body motor patterns appear sufficiently well developed to enable effective esophageal volume clearance of acidic refluxate. We have shown previously in healthy premature infants that transient LES relaxations are the predominant mechanism of GER identified manometrically by common cavity phenomena (abrupt sustained pressurization of the esophageal body)23 and by pH probe.*” In addition, the values for the frequency of GER and transient LES relaxation, esophageal pressure wave sequences during GER episodes, duration of GER episodes as well as of LES relaxation parameters in preterm infants with CLD are in general agreement with our previously reported values in healthy preterm infants.23z24Hence, these data indicate that the mechanisms responsible for competence of the esophagogastric pressure barrier (ie, LES resting tone), associated with the triggering of GER (ie, transient LES relaxation rate) and responsible for esophageal volume clearance (ie, peristaltic esophageal pressure wave sequences) are no different from those in healthy premature control infants.‘3,24%27 Methylxanthines (caffeine and theophylline) have been shown previously to augment GER in premature infants with idiopathic apnea,29 an effect possibly caused by augmented gastric acid secretion. In our study, the presence of theophylline-caffeine therapy did not significantly alter the frequency of acid GER. This observation contrasts with that in our previous study of a larger population of premature infants who did not have CLD, which suggested that theophylline also may augment the frequency of transient LES relaxations. However, because our study was not designed specifically to examine the impact of methylxanthine treatment and the total number of infants studied was small, the possible impact

1798

OMARI

of methylxanthines on GER requires further investigation. Results of the current study have shown that in premature infants with CLD, as in healthy premature infants, older children, and adults, the major mechanism underlying GER is transient LES relaxation. In addition, the patterns of reflux and motor mechanisms responsible

ET AL

for esophageal volume clearance are similar to those in premature infants without CLD.

ACKNOWLEDGMENT The authors thank Dr Charles Malbert for the computerized acquisition and analysis software used in this study.

data

REFERENCES 1. Hrabovsky E, Mullett M: Gastroesophageal reflux and the premature infant. J Pediatr Surg 21:583-587, 1986 2. Guiffre R, Rubin S, Mitchell I: Antireflux surgery in infants with bronchopulmonary dysplasia. Am J Dis Child 141:648-651, 1987 3. Cyr J, Ferrara T, Thompson T, et al: Treatment of pulmonary manifestations of gastroesophageal reflux in children two years of age or less. Am J Surg 157:400-404, 1989 4. Foglia R, Fonkalsrud E, Ament M, et al: Gastroesophageal fundoplication for the management of chronic pulmonary disease in children. Am .I Surg 140:72-79, 1980 5. Herbst J, Minton S, Book L: Gastroesophageal reflux causing respiratory distress and apnea in newborn infants. .I Pediatrics 95:763768,1979 6. Jolley S, Halpern C, Sterling C, et al: The relationship of respiratory complications from gastroesophageal reflux to prematurity in infants. .I Pediatr Surg 25:755-757,199O 7. Sindel B, Maisels M, Ballantine T: Gastroesophageal reflux to the proximal esophagus in infants with bronchopulmonary dysplasia. Am J Dis Child 143:1103-1106, 1989 8. Orenstein S, Orenstein D: Gastroesophageal reflux and respiratory disease in children. J Pediatrics 112:847-858.1988 9. Forster L, Trudeau W, Goldman A: Bronchodilator effects on gastric acid secretion. JAMA 241:2613-2615, 1979 10. Stein M, Towner T, Weber R, et al: The effect of theophylline on the lower esophageal sphincter pressure. Ann Allergy 45:238-241,198O 11. Berquist W, Rachelefsky G, Kadden M, et al: Effect of theophylline on gastroesophageal reflux in normal adults. J Allergy Clin Immunol67:407-411, 1981 12. Newell S, Morgan M, Durbin G, et al: Does mechanical ventilation precipitate gastro-oesophageal reflux during enteral feeding. Arch Dis Child 64:1352-1355, 1989 13. Cameron B, Co&ran W, McGill C: The uncut Collis-Nissen fundoplication: Results for 79 consecutively treated high risk children. J Pediatr Surg 32:887-891,1997 14. Mansfield L: Gastro-oesophageal reflux and respiratory disorders: A review. Ann Allergy 62: 158, 1989 15. Christie D, O’Grady L, Mack D: Incompetent lower esophageal sphincter and gastroesophageal reflux in recurrent acute pulmonary disease of infancy and childhood. J Pediatrics 93:23-27, 1978

16. Werlin S, Dodds W, Hogan W, et al: Mechanisms of gastroesophageal reflux in children. J Pediatrics 97:244-249, 1980 17. Kawabara H, Dent J, Davidson G: Evaluation of the mechanisms responsible for gastroesophageal reflux in children. Gastroenterology 113:399-408,1997 IS. Mittal R, Holloway R, Penagini R, et al: Transient lower esophageal relaxation. Gastroenterology 109:601-610, 1995 19. Cucchiara S, Bortolotti M, Minella R, et ah Fasting and postprandial mechanisms of gastroesophageal reflux in children with gastroesophageal reflux. Dig Dis Sci 38:86-92, 1993 20. Dent I, Dodds W, Friedman R, et al: Mechanism of gastroesophageal reflux in recumbent asymptomatic human subjects. J Clin Invest 65:256-267, 1980 21. Dent J, Holloway R, Toouli J. et al: Mechanisms of lower oesophageal sphincter incompetence in patients with symptomatic gastro-oesophageal reflux. Gut 29:1020-1028,1988 22. Holloway R, Penagini R, Ireland A: Criteria for the objective definition of transient lower esophageal sphincter relaxation. Am J Physiol268:Gl-8, 1995 23. Omari T, Miki K, Davidson G, et al: Characterisation of relaxation of the lower oesophageal spnicter in healthy premature infants. Gut 40:370-375. 1997 24. Omari T, Barnett C, Snel A, et al: Mechanisms of gastroesophageal reflux in healthy premature infants. J Pediatrics 133:650-654, 1998 25. Northway WH, Rosan RC, Porter DY: Pulmonary disease following respiratory therapy of hyaline membrane disease: BPD. N Engl J Med 276:357. 1967 26. Tote S, Farrell P. Leavitt L, et al: Clinical and roentgenographic scoring systems for assessing bronchopulmonary dysplasia. Am J Dis Child 138:581-585, 1984 27. Omari T, Miki K, Fraser R, et al: Esophageal body and lower esophageal sphincter function in healthy premature infants. Gastroenterology 109:1757-1764,1995 28. Gmari T, Bakewell M, Fraser R, et al: Intraluminal micromanometry: An evaluation of the dynamic perfomlance of micro-extrusions and sleeve sensors. Neurogastroenterol Motil8:241-245,1996 29. Sacre L, Vandenplas Y Xanthines in apnea of premature infants. Influence on gastroesophageal reflux. Arch Fr Pediatr 44:383-385,1987