November 1979
The Journal o f P E D I A T R I C S
763
Gastroesophageal reflux causing respiratory distress and apnea in newborn infants Respiratory distress, apnea, and chronic puhTwnar), disease since birth were identified in 14 infants who also had symptomatic gastroesophageal reflux. Birth weights varied from 760 to 4,540 gm. All infants had radiographic changes similar to those in bronchopulmona O' dysplasia. Cessation of apnea and improvement of pulmonary disease occurred only after medical (8) or surgical (6) control of gastroesophageal reflnx. Simultaneous tracings of esophageal pH, heart rate, impedance pneumography, and nasal air flow in ,five inJants demonstrated that reflnx preceded apnea. Apnea could be induced by instillation of dihtte acid, but not water or formula, into the esophagus. Prolonged monitoring of esophagea[ p H more than two hours after feeding in 14 other infants less than 6 weeks of age (birth weight 780 to 3,350 gm) without a history of recent vomiting indicated that reflux was not greater than in normal older children.
John J. Herbst, M.D.,* Stephen D. Minton, M.D., and Linda S. Book, M.D., S a l t L a k e City, U t a h
S YMP 1-o ~ s of respiratory distress, apnea, bradycardia, and cyanosis are common in newborn infants, especially premature infants. ~-4 These symptoms often persist in infants who develop chronic pulmonary disease? ~ Clinical evaluation of these patients suggested that in some instances the course or progression of the lung disease may be augmented by aspiration of refluxed gastric contents. The authors have identified 14 infants with respiratory distress starting at birth in whom chronic gastroesophageal reflux contributed to chronic lung disease, apnea, and bradycardia.
CASE MATERIAL AND M E T H O D S Fourteen infants were identified who had been hospitalized at the University of Utah Medical Center since birth with respiratory distress and recurrent apnea. An apneic episode was defined as cessation of breathing in excess of 20 seconds. All infants had multiple episodes of apnea, and assisted ventilation with a bag and mask and
From the Department of Pediatrics, University of Utah School of Medicine. Sttpported in part by a grant.fi'om the Thrasher Research Fund. *Reprint address: Department of Pediatrics, Universi(vof Utah Medical Centen Salt Lake Cio; UT 84132.
0022-3476/79/110763 + 06500.60/0 9 1979 The C. V. Mosby Co.
use of increased FI% were frequently necessary. These 14 infants were identified over a four-year period, during which there were 2,418 admissions to the newborn intensive care unit, Five of the 14 infants were term and had a mean birth weight of 3,048 gin; nine were premature, with a mean birth weight of 1,389 gm and a mean gestational age of 31 weeks. Abbreviation used FIo: fraction inspiratory oxygen These 14 patients were included in the present series because it could be demonstrated that gastroesophageal reflux with aspiration was a major factor affecting the severity of the chronic lung disease. Clinical suspicion of gastroesophageal reflux was usually aroused by the presence of gastric contents (curdled milk or bilious material) in aspirates of the pharynx or trachea during apneic episodes, hematemesis, frequent emesis, or recurrent episodes of atelectasis as noted on chest radiographs. Clinical suspicion of abnormal gastroesophageal reflux was first noted at 2 to 24 weeks of life. Frequent reflux was noted on barium swallow in all 14 patients. An esophageal pH test (Tuttle) was performed in 13 infants ~ s: 300 ml/1.7 m-' of 0.1 normal HC1 was instilled into the stomach, and the distal esophageal pH 3 cm
Vol. 95, No. 5, part 1, pp. 763-768
Herbst, Minton, and Book
764
The Journal of Pediatrics November 1979
Table. Respiratory symptoms with reflux
Patient
Birth weight (gin)
1 2 3 4 5 6 7 8 9 10 11
4,540 2,940 2,650 2,560 2,550 2,400 1,956 1,550 980 1,650 1,160
12 13 14
1,160 885 760
Perinatal history ASP/BA ASP RDS/BA RDS/BA RDS RDS RDS/PFS RDS RDS RDS/BA RDS/PFS/ BA RDS/BA RPS/BA PDS/BA
Early respiratory support
Apneic episodes
% F1%
Therapy
I
Day 0 I Day 7
Day 0
Day 7
Type
Age (wk)
CPAP CPAP IPPB/PEEP CPAP CPAP CPAP IPPB/PEEP IPPB/PEEP IPPB/PEEP IPPB/PEEP IPPB/PEEP
1/day 1/ day 4/day 3/day 4/day 13/day 2/day 4/day .4/day 8/day 4/day
0 0 1/wk 0 0 0 1/wk 0 2/wk 0 2/wk
0.25 0.27 0.25 0.25 0.20 0.23 0.25 0.25 IPPB 0.24 0.35 + IPPB 0.20
0,20 0.20 0.22 0.20 0.20 0.20 0.20 0.21 0.20 0.20 0.20
Medical Medical Medical Surgical Medical Surgical Surgical Medical Medical Surgical Medical
2 12 24 8 2 12 22 4 5 6 11
IPPB/PEEP IPPB/PEEP IPPD/PEEP
3/day 20/day 8/day
0 1/day 1/wk
0.22 0.30 30 IPPB
0.20 0.23 0.20
Surgical Medical Surgical
9 7 9
All patients had radiographs that were compatible with bronchopulmonary dysplasia. RDS = Respiratory distress syndrome; ASP = aspiration pneumonia at birth; BA = birth asphyxia; PFS = persistent fetal circulation; CPAP = continous positive airway pressure; IPPB/PEEP = Intermittent positive pressure breathing with positive end-expiratory pressure. Day 0 = start of therapy; Day 7 = 1 week later. See text for further details.
proximal to the gastroesophageal junction was recorded. The pH probe was standardized before and after use, and the tracings were not used if standardization at pH 4 drifted more than 0.1 pH units. An abnormal test was defined as at least two episodes of reflux in 20 minutes, detected by a decrease in esophageal pH to less than 4 for more than 15 seconds. All infants received intensive medical treatment for gastroesophageal reflux for at least three weeks and. in one instance, this treatment was continued for 16 weeks. Medical management included frequent small-volume feedings with the child semiupright or on the right side. and trials of antacid agents. Constant nasogastric gavage, nasojejunal feedings, intravenous alimentation, and cessatiofi of all oral feedings were also employed in various patients. In five patients, simultaneous recordings of esophageal pH 3 cm above the lower esophageal sphincter, impedance pneumogram, electrocardiogram, and tracing of nasal air flow were obtained to determine the sequence of the events of the apnea-bradycardia-gastroesophageal reflux episodes. Esophageal pH was monitored with a 1.5 mm flexible pH probe (Microelectrodes, Inc.. Londonderry, N.H.). A Tektronics neonatal monitor was used to monitor impedance pneumography and the electrocardiogram. A nasal thermistor was constructed by the Department of Bioengineering at the Un!yersity of Utah to measure nasal air flow. Data from all of the equipment were simultaneously recorded on a Gilson multichannel recorder.
In order to determine if episodes of reflux are normally more frequent in the first weeks of life, prolonged monitoring of esophageal pH was performed in 14 other hospitalized infants less than 6 weeks of age. Their birth weights varied from 780 to 3,350 gin. N o n e of the infants had a recent history of vomiting, but five were having recurrent episodes of idiopathic apnea. I n f o r m e d consent of the parents and lack of a recent history of vomiting were the only criteria used in patient selection. O f these, the ten infants weighing less than 2,500 gm at birth required mechanical ventilatory assistance with increased FI,,, at some time. Esophageal pH 3 cm above the lower esophageal sphincter was recorded for 18 to 24 hours, and the episodes of reflux m o r e than two hours after feeding were scored. T,l~e p r e s e n c e of gastric acidity was confirmed in each patient at the time the pH probe was inserted. Previous experience in older children and adults has shown this to be a very sensitive quantitative test for presence of gastroesophageal reflux which clearly identifies patients with symptomatic refluxY A numerical score was assigned to the study based on the f611o~ing criteria: (1) number of episodes of reflux greater than 15 seconds, (2) reflux episodes longer than five minutes, (3) longest reflux episode, and (4) percent of time with pH less than 4. These criteria were applied to the tracings obtained while the child was awake, asleep, sitting, and lying supine. Thus. 16 subscores were obtained. Based on studies in normal older children, 2 points were assigned to the m e a n of each subscore, and
Volume 95 Number 5, part 1
Gastroesophageal reflux and apnea
another point was added for each standard deviation above the mean. r' The scores from these 14 infants without clinical reflux were compared to the scores of the previously reported older children using the Student t test.
765
RR 38/MIN
RESULTS All 14 of the symptomatic infants had moderate or severe gastroesophageal reflux noted at barium swallow, and a positive esophageal pH test was noted in each of the 13 infants tested. Severe inflammation and easy friability of the esophagus were present in five patients at the time of surgery, and four patients had frank hematemesis. All 14 infants had chest radiographs showing areas of streaky atelectasis and areas of hyperlucency, interpreted as being compatible with bronchopulmonary dysplasia. These changes were noted by 4 weeks of age in all patients. Episodes of apnea were noted in the first week of life in all patients. The birth weight, pertinent diagnoses in the first week of life, and the most aggressive respiratory support provided in the first week of life are presented in the Table. Intensive medical therapy of the documented gastroesophageal reflux led to a prompt decrease and ultimate cessation of apneic spells in four of the five term infants and in four of the nine premature infants. In these eight patients, improvement was directly related to improvement in gastroesophageal reflux and cessation of aspiration, and was not due to healing of bronchopulmonary dysplasia. Prior to institution of anti-reflux treatment. the clinical course of these patients had either been stable or gradually had become worse over at least a two-week period. In successfully treated patients, gastric contents were no longer present in aspirated pharyngeal or tracheal fluid, apnea decreased, and there was a rapid decrease in need for oxygen. Steady clinical improvement continued, with cessation of all apneic episodes. No new infiltrates were noted on chest radiographs, and old infiltrates started to clear. The Table shows the number of apneic episodes experienced per day and the FI,,., for the period just before the onset of effective therapy, and again after one week of therapy. In six patients, medical therapy was unsuccessful, but there was dramatic improvement after successful surgery to prevent gastroesophageal reflux. For these patients, the number of apneic episodes and the FI,,., in the Table refer to the period just prior to surgery and again one week after surgery. Three patients were successfully removed from mechanical ventilators within a week following surgery, after previous attempts to do so over at least a three-week period had been unsuccessful. In all five patients (10 through 14. Table) who were extensively studied to document the sequence of events in
HR 150/MIN
76-
pH 5" 432I-
I0 SEC
2--
Fig. 1. Tracing of vital signs of a patient demonstrating that apnea follows a decrease in esophageal pH. From top to bottom the tracings record the impedance pneumotachogram, tracing of thermistor measuring air flow at the nostril, electrocardiogram, and esophageal pH.
the apneic episodes, it was demonstrated that apnea occurred immediately after gastroesophageal reflux and in some cases was associated with upper airway obstruction. The tracings of these five patients were all very similar. Fig. 1 shows tracings of heart rate, esophageal pH, impedance pneumography, and air movement at the nostril from a representative patient. An episode of spontaneous reflux of acidic gastric contents was followed by 24 seconds of apnea with no air movement detected at the nostril, while the impedance pneumotachograph showed evidence of movement of the chest wall. These tracings recorded what clinically appeared to be laryngospasm. Spontaneous reflux-induced apnea occurred only three to four times per day in two infants, and obtaining a satisfactory recording of all four tracings at the time of an apneic episode was difficult. A recording demonstrating the sequence of an acid esophageal pH followed by apnea could be easily obtained in all infants by simulating reflux, as shown in Fig. 2. This was done by instilling, over a 15-second period of time, 0.5 ml of 0.1N HC1 into the midesophagus via a 6 Fr. feeding tube. In addition to the apnea noted in all five patients, there was a decrease in heart rate of at least 30 beats per minute for at least 20
766
Herbst, Minton, and Book
The Journal of Pediatrics November 1979
A.
difficult to maintain the infants in an upright position in an infant incubator. Although a few infants had mildly elevated scores (normal = < 16), the mean scores were not Significantly greater than those obtained in older infants. In the six control infants with idiopathic apnea, there was no temporal relationship of apnea and spontaneous or induced reflux.
RR
HR 138
DISCUSSION
I / 2 ml HCL
76 pH 5 4.
3-
- ~ ~ _
2l-
B.
7 "-I pH
112 cc MILK /
6
~
~'
5
Fig. 2. The arrangements of recording"as in Fig. 1. A, Apnea induced by instillation of 0. IN HCI into the midesophagus. B, No effect with instillation of milk. seconds in four patients. Instillation of dilute formula or water caused no changes in vital signs in any of the five infants studied. Similar instillations of 0.1N HCI, water or formula in six infants without reflux caused no change in heart or respiratory rate. In the 14 infants without a history of symptomatic reflux, calculation of the numerical scores after prolonged monitoring of esophageal pH were ~&&ollo~vs: awake. 12.6+ 2.2 SEM; asleep 11.4 _+ 1.4:~upine 14.6 • 2,4: upright 10.8 • 1.9. The upright score could only be calculated in six of the 14 infants because it was very
Newborn infants with respiratory distress requiring assisted mechanical ventilation and high levels of inspired oxygen frequently have apnea, bradycardia, and cyanosis?...... Many of these infants subsequently develop bronchopulmonary dysplasia.:'. ~ We observed a group of infants in whom gastroesophageal reflux and aspiration were major factors in the development of chronic lung disease and apnea. The radiographic sequelae of chronic aspiration may be similar to the changes of bronchopulmonary dysplasia. The dramatic improvement in chest radiographs, apneic episodes, and FI .... requirements following control of gastroesophageal reflux suggests that reflux was much more important than bronchopulmonary dysplasia in the evolution of the chronic lung disease in these patients. The premature infants were thought to have bronchopulmonary dysplasia in addition to the sequelae of gastroesophageal reflux. In these 14 patients, the presence of severe esophagitis in five infants, complicated by hematemesis in four, illustrates the severity of the gastroesophageal reflux. Reflux or vomiting is so common in newborn infants that it is often considered to be a variation of normal and is frequently ignored if weight gain continues. Prolonged monitoring of esophageal pH was done to see if frequent reflux of gastric acid into the esophagus is a normal event. Prolonged monitoring for 18 to 24 hours was employed because the length of observation tends to minimize error, owing to the short observation period inherent in a barium swallow ~ d y or acid perfusion test. Since the amount of reflux is expressed as a numerical score, scores from groups of patients can be compared. The 14 asymptomatic infants were all studied in the first eight weeks of life and had an average birth weight of 1,786 gm. The pH scores may be compared with older children, in whom the upper limit for each score is 16 points'." Although some of the neonates had mild elevations in the scores, the mean scores were not statistically different than those obtained in normal older children. Thus the integrity of the lower esophageal sphincter mechanism seems to be well-developed in the first weeks of life, even in very small premature infants. Although some reflux is present in all infants, the reflux noted in the
Volume 95 Number 5, part 1
fourteen symptomatic patients was definitely abnormal. Monitoring of esophageal pH does not distinguish reflux of gastric contents into the esophagus from reflux leading to vomiting of gastric contents from the mouth. The clinically observed frequency of "normal spitting up" in infants may indicate that infants are more likely than older children to propel refluxed material into the mouth. Since infants may vomit in response to a variety of stimuli, it became crucial to determine if the episodes of reflux preceded and presumably induced the apneic spells, or if reflux occurred during apnea as a late event of hypoxia. We determined that recording chest impedance was an incomplete method to monitor respiration. We had noted episodes of laryngospasm, with opposed vocal cords seen at laryngoscopy, in four of the 14 symptomatic infants. In the apnea-reflux episodes in these patients, a decrease in esophageal pH preceded the apnea (Fig. 1). An episode of presumed laryngospasm similar to those noted at laryngoscopy was also indicated by the finding of continued chest movement without air movement at the nostril. Nasopharyngeal obstruction could have caused a similar record, and it is likely that such obstruction could have occurred in some of our patients. Since episodes of laryngospasm were noted at laryngoscopy following similar events in three patients, including the patient whose record is shown in Fig. 1, it is reasonable to conclude that laryngospasm was caused by reflux in at leas( some of the patients. The instillation of 0.5 ml of 0.1 normal HC1 into the midesophagus provided a simple method of demonstrating that the presence Of material with an acid pH in the esophagus could induce apnea. When apneic episodes were not frequent, often one or more of the monitors was not recording events optimally at the time of a spontaneous episode of reflux-induced apnea. The response to acid in the patients with apnea was a specific one, since instillation of water or formula did not evoke apnea and, in asymptomatic patients, instillation of acid caused no change in respiratory pattern of heart rate. The mechanisms for the response are not clear, but there are chemoreceptors about the larynx and pharynx. It has also been shown that instillation of acid but not of water or autologous milk into the larynx of newborn lambs will cause apnea. 1~Dilute solutions of tracheal fluid or saline instilled on the larynx of fetal iambs are also more potent inhibitors of respiration than isotonic solutions. By three months of age, the lamb is able to alternate swallows between breaths without periods of apnea. 1:' Stimulation of the chemoreceptors about the pharynx may have induced the laryngospasm in our infants. Frequent reflux and severe esophagitis may sensitize
Gastroesophageal reflux and apnea
767
chemoreceptors so as to elicit an exaggerated response. In addition, symptoms of heartburn frequently occur in adults without microscopic evidence of esophagitis, 14 suggesting that pain receptors may become sensitized in all the patients with reflux, and not just those with severe esophagitis. Recently, Ariagno and associates ~:' noted periods of apnea associated with high airway obstruction in infants having recurrent apnea during nonrapid eye movement sleep. Unfortunately, no special search was made to detect gastroesophageal reflux. Our group of 14 patients with symptomatic gastroesophageal reflux demonstrates that the sequence of reflux, apnea, laryngospasm, and the development of chronic pulmonary disease may mimic bronchopulmonary dysplasia. Thus, in infants with apnea, laryngospasm, upper airway obstruction, and a typical development of what appears to be bronchopulmonary dysplasia, gastroesophageal reflux must be considered as a possible cause.
REFERENCES
1. Farrell PM, and Avery ME: Hyaline membrane disease, Am Rev Respir Dis 3:657, 1975. 2. Farrell PM, and Wood RE: Epidemiology of hyaline membrane disease in the United States: Analysisof national mortality statistics, Pediatrics 58:167, 1976. 3. Daily W, Klaus M, and Meyer H: Apnea in premature infants, Pediatrics 43:510, 1969. 4. Kanwinkel J, ,Nearman HS, Fanaroff AA, Katona PG, and Klaus MH: Apnea of prematurity, J PEDIATR 86:588, 1975. 5. Watts JL, Ariagno RL, and Brady JP: Chronic pulmonary disease in neonates after artificial ventilation: Distribution of ventilation in pulmonary interstitial emphysema, Pediatrics 60:273, 1977. 6. Klaus M: Respiratory function and puhnonary disease in the newborn, in Barnett H, editor: Pediatrics, New York, 1972, Appleton-Century-Crofts, pp 1255-61. 7. Johnson DG, Herbst JJ, Oliveros MA. and Stewart DR: Evaluation of gastroesophageal reflux surgery in children, Pediatrics 59:62, 1977. 8. Euler AR and Ament ME: Detection of gastroesophageal reflux by tuttle test, Pediatrics 60:65, 1977. 9. Jolley SG, Johnson DG, Herbst JJ, Pena RA, and Garnier R: An assessment of gastroesophageal reflux in children by extended pH monitoring of the distal esophagus, Surgery 84:16, 1978. 10. Schafl'erAJ and Avery ME: Disease of the newborn, ed 4, Philadelphia, 1977, WB Saunders Company, p 315. 11. Avery GB, and Fletcher AB: Nutrition, in Avery GB: Neonatology pathophysiology and management of the newborn, Philadelphia, 1975, JB Lippincon Company, p 878. 12. Johnson P: Laryngeal induced apnea, in Robinson RR. editor: SIDS 1974, proceeding of the Francis E. Camps international symposium on sudden and unexpected deaths
768
13.
14.
Herbst, Minton, and Book
The Journal of Pediatrics November 1979
in infancy, Canada, 1974, The Canadian Foundation for the Study Of Infant Death, p 878. Harned HS Jr, Myracle J and Ferreiro J: Respiratory suppression and swallowing from introduction of fluids into the laryngeal region of the lamb, Pediatr Res 12:1003, 1978. Behar J, Biancani P and Sheahan DG: Evaluation of
15.
esophageal tests in the diagnosis of reflux esophagitis, Gastroenterology 71:9, 1976. Ariagno RL, Guilleminault C, and Negel LE: Mixed and obstructive sleep apnea in 3-month-old control and near miss for sudden infant death syndrome infants, Pediatr Res 12:519. 1978.
Copyright information The appearance of a code at the bottom of the first page of an original article in this JOURNAL indicates the copyright owner's consent that copies of the article may be made for personal or internal use, or li)r the personal or internal use of specific clien;t~:.SThisconsent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc., P.O. BOx 765, Schenectady, N.Y. 12301, (518) 374-4430, for copyin~ beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying l~or general distribution, for advertising or promotional purposes, for creating new collective works, or for resale.
The Journal o f P E D I A T R I C S
763
Gastroesophageal reflux causing respiratory distress and apnea in newborn infants Respiratory distress, apnea, and chronic puhTwnar), disease since birth were identified in 14 infants who also had symptomatic gastroesophageal reflux. Birth weights varied from 760 to 4,540 gm. All infants had radiographic changes similar to those in bronchopulmona O' dysplasia. Cessation of apnea and improvement of pulmonary disease occurred only after medical (8) or surgical (6) control of gastroesophageal reflnx. Simultaneous tracings of esophageal pH, heart rate, impedance pneumography, and nasal air flow in ,five inJants demonstrated that reflnx preceded apnea. Apnea could be induced by instillation of dihtte acid, but not water or formula, into the esophagus. Prolonged monitoring of esophagea[ p H more than two hours after feeding in 14 other infants less than 6 weeks of age (birth weight 780 to 3,350 gm) without a history of recent vomiting indicated that reflux was not greater than in normal older children.
John J. Herbst, M.D.,* Stephen D. Minton, M.D., and Linda S. Book, M.D., S a l t L a k e City, U t a h
S YMP 1-o ~ s of respiratory distress, apnea, bradycardia, and cyanosis are common in newborn infants, especially premature infants. ~-4 These symptoms often persist in infants who develop chronic pulmonary disease? ~ Clinical evaluation of these patients suggested that in some instances the course or progression of the lung disease may be augmented by aspiration of refluxed gastric contents. The authors have identified 14 infants with respiratory distress starting at birth in whom chronic gastroesophageal reflux contributed to chronic lung disease, apnea, and bradycardia.
CASE MATERIAL AND M E T H O D S Fourteen infants were identified who had been hospitalized at the University of Utah Medical Center since birth with respiratory distress and recurrent apnea. An apneic episode was defined as cessation of breathing in excess of 20 seconds. All infants had multiple episodes of apnea, and assisted ventilation with a bag and mask and
From the Department of Pediatrics, University of Utah School of Medicine. Sttpported in part by a grant.fi'om the Thrasher Research Fund. *Reprint address: Department of Pediatrics, Universi(vof Utah Medical Centen Salt Lake Cio; UT 84132.
0022-3476/79/110763 + 06500.60/0 9 1979 The C. V. Mosby Co.
use of increased FI% were frequently necessary. These 14 infants were identified over a four-year period, during which there were 2,418 admissions to the newborn intensive care unit, Five of the 14 infants were term and had a mean birth weight of 3,048 gin; nine were premature, with a mean birth weight of 1,389 gm and a mean gestational age of 31 weeks. Abbreviation used FIo: fraction inspiratory oxygen These 14 patients were included in the present series because it could be demonstrated that gastroesophageal reflux with aspiration was a major factor affecting the severity of the chronic lung disease. Clinical suspicion of gastroesophageal reflux was usually aroused by the presence of gastric contents (curdled milk or bilious material) in aspirates of the pharynx or trachea during apneic episodes, hematemesis, frequent emesis, or recurrent episodes of atelectasis as noted on chest radiographs. Clinical suspicion of abnormal gastroesophageal reflux was first noted at 2 to 24 weeks of life. Frequent reflux was noted on barium swallow in all 14 patients. An esophageal pH test (Tuttle) was performed in 13 infants ~ s: 300 ml/1.7 m-' of 0.1 normal HC1 was instilled into the stomach, and the distal esophageal pH 3 cm
Vol. 95, No. 5, part 1, pp. 763-768
Herbst, Minton, and Book
764
The Journal of Pediatrics November 1979
Table. Respiratory symptoms with reflux
Patient
Birth weight (gin)
1 2 3 4 5 6 7 8 9 10 11
4,540 2,940 2,650 2,560 2,550 2,400 1,956 1,550 980 1,650 1,160
12 13 14
1,160 885 760
Perinatal history ASP/BA ASP RDS/BA RDS/BA RDS RDS RDS/PFS RDS RDS RDS/BA RDS/PFS/ BA RDS/BA RPS/BA PDS/BA
Early respiratory support
Apneic episodes
% F1%
Therapy
I
Day 0 I Day 7
Day 0
Day 7
Type
Age (wk)
CPAP CPAP IPPB/PEEP CPAP CPAP CPAP IPPB/PEEP IPPB/PEEP IPPB/PEEP IPPB/PEEP IPPB/PEEP
1/day 1/ day 4/day 3/day 4/day 13/day 2/day 4/day .4/day 8/day 4/day
0 0 1/wk 0 0 0 1/wk 0 2/wk 0 2/wk
0.25 0.27 0.25 0.25 0.20 0.23 0.25 0.25 IPPB 0.24 0.35 + IPPB 0.20
0,20 0.20 0.22 0.20 0.20 0.20 0.20 0.21 0.20 0.20 0.20
Medical Medical Medical Surgical Medical Surgical Surgical Medical Medical Surgical Medical
2 12 24 8 2 12 22 4 5 6 11
IPPB/PEEP IPPB/PEEP IPPD/PEEP
3/day 20/day 8/day
0 1/day 1/wk
0.22 0.30 30 IPPB
0.20 0.23 0.20
Surgical Medical Surgical
9 7 9
All patients had radiographs that were compatible with bronchopulmonary dysplasia. RDS = Respiratory distress syndrome; ASP = aspiration pneumonia at birth; BA = birth asphyxia; PFS = persistent fetal circulation; CPAP = continous positive airway pressure; IPPB/PEEP = Intermittent positive pressure breathing with positive end-expiratory pressure. Day 0 = start of therapy; Day 7 = 1 week later. See text for further details.
proximal to the gastroesophageal junction was recorded. The pH probe was standardized before and after use, and the tracings were not used if standardization at pH 4 drifted more than 0.1 pH units. An abnormal test was defined as at least two episodes of reflux in 20 minutes, detected by a decrease in esophageal pH to less than 4 for more than 15 seconds. All infants received intensive medical treatment for gastroesophageal reflux for at least three weeks and. in one instance, this treatment was continued for 16 weeks. Medical management included frequent small-volume feedings with the child semiupright or on the right side. and trials of antacid agents. Constant nasogastric gavage, nasojejunal feedings, intravenous alimentation, and cessatiofi of all oral feedings were also employed in various patients. In five patients, simultaneous recordings of esophageal pH 3 cm above the lower esophageal sphincter, impedance pneumogram, electrocardiogram, and tracing of nasal air flow were obtained to determine the sequence of the events of the apnea-bradycardia-gastroesophageal reflux episodes. Esophageal pH was monitored with a 1.5 mm flexible pH probe (Microelectrodes, Inc.. Londonderry, N.H.). A Tektronics neonatal monitor was used to monitor impedance pneumography and the electrocardiogram. A nasal thermistor was constructed by the Department of Bioengineering at the Un!yersity of Utah to measure nasal air flow. Data from all of the equipment were simultaneously recorded on a Gilson multichannel recorder.
In order to determine if episodes of reflux are normally more frequent in the first weeks of life, prolonged monitoring of esophageal pH was performed in 14 other hospitalized infants less than 6 weeks of age. Their birth weights varied from 780 to 3,350 gin. N o n e of the infants had a recent history of vomiting, but five were having recurrent episodes of idiopathic apnea. I n f o r m e d consent of the parents and lack of a recent history of vomiting were the only criteria used in patient selection. O f these, the ten infants weighing less than 2,500 gm at birth required mechanical ventilatory assistance with increased FI,,, at some time. Esophageal pH 3 cm above the lower esophageal sphincter was recorded for 18 to 24 hours, and the episodes of reflux m o r e than two hours after feeding were scored. T,l~e p r e s e n c e of gastric acidity was confirmed in each patient at the time the pH probe was inserted. Previous experience in older children and adults has shown this to be a very sensitive quantitative test for presence of gastroesophageal reflux which clearly identifies patients with symptomatic refluxY A numerical score was assigned to the study based on the f611o~ing criteria: (1) number of episodes of reflux greater than 15 seconds, (2) reflux episodes longer than five minutes, (3) longest reflux episode, and (4) percent of time with pH less than 4. These criteria were applied to the tracings obtained while the child was awake, asleep, sitting, and lying supine. Thus. 16 subscores were obtained. Based on studies in normal older children, 2 points were assigned to the m e a n of each subscore, and
Volume 95 Number 5, part 1
Gastroesophageal reflux and apnea
another point was added for each standard deviation above the mean. r' The scores from these 14 infants without clinical reflux were compared to the scores of the previously reported older children using the Student t test.
765
RR 38/MIN
RESULTS All 14 of the symptomatic infants had moderate or severe gastroesophageal reflux noted at barium swallow, and a positive esophageal pH test was noted in each of the 13 infants tested. Severe inflammation and easy friability of the esophagus were present in five patients at the time of surgery, and four patients had frank hematemesis. All 14 infants had chest radiographs showing areas of streaky atelectasis and areas of hyperlucency, interpreted as being compatible with bronchopulmonary dysplasia. These changes were noted by 4 weeks of age in all patients. Episodes of apnea were noted in the first week of life in all patients. The birth weight, pertinent diagnoses in the first week of life, and the most aggressive respiratory support provided in the first week of life are presented in the Table. Intensive medical therapy of the documented gastroesophageal reflux led to a prompt decrease and ultimate cessation of apneic spells in four of the five term infants and in four of the nine premature infants. In these eight patients, improvement was directly related to improvement in gastroesophageal reflux and cessation of aspiration, and was not due to healing of bronchopulmonary dysplasia. Prior to institution of anti-reflux treatment. the clinical course of these patients had either been stable or gradually had become worse over at least a two-week period. In successfully treated patients, gastric contents were no longer present in aspirated pharyngeal or tracheal fluid, apnea decreased, and there was a rapid decrease in need for oxygen. Steady clinical improvement continued, with cessation of all apneic episodes. No new infiltrates were noted on chest radiographs, and old infiltrates started to clear. The Table shows the number of apneic episodes experienced per day and the FI,,., for the period just before the onset of effective therapy, and again after one week of therapy. In six patients, medical therapy was unsuccessful, but there was dramatic improvement after successful surgery to prevent gastroesophageal reflux. For these patients, the number of apneic episodes and the FI,,., in the Table refer to the period just prior to surgery and again one week after surgery. Three patients were successfully removed from mechanical ventilators within a week following surgery, after previous attempts to do so over at least a three-week period had been unsuccessful. In all five patients (10 through 14. Table) who were extensively studied to document the sequence of events in
HR 150/MIN
76-
pH 5" 432I-
I0 SEC
2--
Fig. 1. Tracing of vital signs of a patient demonstrating that apnea follows a decrease in esophageal pH. From top to bottom the tracings record the impedance pneumotachogram, tracing of thermistor measuring air flow at the nostril, electrocardiogram, and esophageal pH.
the apneic episodes, it was demonstrated that apnea occurred immediately after gastroesophageal reflux and in some cases was associated with upper airway obstruction. The tracings of these five patients were all very similar. Fig. 1 shows tracings of heart rate, esophageal pH, impedance pneumography, and air movement at the nostril from a representative patient. An episode of spontaneous reflux of acidic gastric contents was followed by 24 seconds of apnea with no air movement detected at the nostril, while the impedance pneumotachograph showed evidence of movement of the chest wall. These tracings recorded what clinically appeared to be laryngospasm. Spontaneous reflux-induced apnea occurred only three to four times per day in two infants, and obtaining a satisfactory recording of all four tracings at the time of an apneic episode was difficult. A recording demonstrating the sequence of an acid esophageal pH followed by apnea could be easily obtained in all infants by simulating reflux, as shown in Fig. 2. This was done by instilling, over a 15-second period of time, 0.5 ml of 0.1N HC1 into the midesophagus via a 6 Fr. feeding tube. In addition to the apnea noted in all five patients, there was a decrease in heart rate of at least 30 beats per minute for at least 20
766
Herbst, Minton, and Book
The Journal of Pediatrics November 1979
A.
difficult to maintain the infants in an upright position in an infant incubator. Although a few infants had mildly elevated scores (normal = < 16), the mean scores were not Significantly greater than those obtained in older infants. In the six control infants with idiopathic apnea, there was no temporal relationship of apnea and spontaneous or induced reflux.
RR
HR 138
DISCUSSION
I / 2 ml HCL
76 pH 5 4.
3-
- ~ ~ _
2l-
B.
7 "-I pH
112 cc MILK /
6
~
~'
5
Fig. 2. The arrangements of recording"as in Fig. 1. A, Apnea induced by instillation of 0. IN HCI into the midesophagus. B, No effect with instillation of milk. seconds in four patients. Instillation of dilute formula or water caused no changes in vital signs in any of the five infants studied. Similar instillations of 0.1N HCI, water or formula in six infants without reflux caused no change in heart or respiratory rate. In the 14 infants without a history of symptomatic reflux, calculation of the numerical scores after prolonged monitoring of esophageal pH were ~&&ollo~vs: awake. 12.6+ 2.2 SEM; asleep 11.4 _+ 1.4:~upine 14.6 • 2,4: upright 10.8 • 1.9. The upright score could only be calculated in six of the 14 infants because it was very
Newborn infants with respiratory distress requiring assisted mechanical ventilation and high levels of inspired oxygen frequently have apnea, bradycardia, and cyanosis?...... Many of these infants subsequently develop bronchopulmonary dysplasia.:'. ~ We observed a group of infants in whom gastroesophageal reflux and aspiration were major factors in the development of chronic lung disease and apnea. The radiographic sequelae of chronic aspiration may be similar to the changes of bronchopulmonary dysplasia. The dramatic improvement in chest radiographs, apneic episodes, and FI .... requirements following control of gastroesophageal reflux suggests that reflux was much more important than bronchopulmonary dysplasia in the evolution of the chronic lung disease in these patients. The premature infants were thought to have bronchopulmonary dysplasia in addition to the sequelae of gastroesophageal reflux. In these 14 patients, the presence of severe esophagitis in five infants, complicated by hematemesis in four, illustrates the severity of the gastroesophageal reflux. Reflux or vomiting is so common in newborn infants that it is often considered to be a variation of normal and is frequently ignored if weight gain continues. Prolonged monitoring of esophageal pH was done to see if frequent reflux of gastric acid into the esophagus is a normal event. Prolonged monitoring for 18 to 24 hours was employed because the length of observation tends to minimize error, owing to the short observation period inherent in a barium swallow ~ d y or acid perfusion test. Since the amount of reflux is expressed as a numerical score, scores from groups of patients can be compared. The 14 asymptomatic infants were all studied in the first eight weeks of life and had an average birth weight of 1,786 gm. The pH scores may be compared with older children, in whom the upper limit for each score is 16 points'." Although some of the neonates had mild elevations in the scores, the mean scores were not statistically different than those obtained in normal older children. Thus the integrity of the lower esophageal sphincter mechanism seems to be well-developed in the first weeks of life, even in very small premature infants. Although some reflux is present in all infants, the reflux noted in the
Volume 95 Number 5, part 1
fourteen symptomatic patients was definitely abnormal. Monitoring of esophageal pH does not distinguish reflux of gastric contents into the esophagus from reflux leading to vomiting of gastric contents from the mouth. The clinically observed frequency of "normal spitting up" in infants may indicate that infants are more likely than older children to propel refluxed material into the mouth. Since infants may vomit in response to a variety of stimuli, it became crucial to determine if the episodes of reflux preceded and presumably induced the apneic spells, or if reflux occurred during apnea as a late event of hypoxia. We determined that recording chest impedance was an incomplete method to monitor respiration. We had noted episodes of laryngospasm, with opposed vocal cords seen at laryngoscopy, in four of the 14 symptomatic infants. In the apnea-reflux episodes in these patients, a decrease in esophageal pH preceded the apnea (Fig. 1). An episode of presumed laryngospasm similar to those noted at laryngoscopy was also indicated by the finding of continued chest movement without air movement at the nostril. Nasopharyngeal obstruction could have caused a similar record, and it is likely that such obstruction could have occurred in some of our patients. Since episodes of laryngospasm were noted at laryngoscopy following similar events in three patients, including the patient whose record is shown in Fig. 1, it is reasonable to conclude that laryngospasm was caused by reflux in at leas( some of the patients. The instillation of 0.5 ml of 0.1 normal HC1 into the midesophagus provided a simple method of demonstrating that the presence Of material with an acid pH in the esophagus could induce apnea. When apneic episodes were not frequent, often one or more of the monitors was not recording events optimally at the time of a spontaneous episode of reflux-induced apnea. The response to acid in the patients with apnea was a specific one, since instillation of water or formula did not evoke apnea and, in asymptomatic patients, instillation of acid caused no change in respiratory pattern of heart rate. The mechanisms for the response are not clear, but there are chemoreceptors about the larynx and pharynx. It has also been shown that instillation of acid but not of water or autologous milk into the larynx of newborn lambs will cause apnea. 1~Dilute solutions of tracheal fluid or saline instilled on the larynx of fetal iambs are also more potent inhibitors of respiration than isotonic solutions. By three months of age, the lamb is able to alternate swallows between breaths without periods of apnea. 1:' Stimulation of the chemoreceptors about the pharynx may have induced the laryngospasm in our infants. Frequent reflux and severe esophagitis may sensitize
Gastroesophageal reflux and apnea
767
chemoreceptors so as to elicit an exaggerated response. In addition, symptoms of heartburn frequently occur in adults without microscopic evidence of esophagitis, 14 suggesting that pain receptors may become sensitized in all the patients with reflux, and not just those with severe esophagitis. Recently, Ariagno and associates ~:' noted periods of apnea associated with high airway obstruction in infants having recurrent apnea during nonrapid eye movement sleep. Unfortunately, no special search was made to detect gastroesophageal reflux. Our group of 14 patients with symptomatic gastroesophageal reflux demonstrates that the sequence of reflux, apnea, laryngospasm, and the development of chronic pulmonary disease may mimic bronchopulmonary dysplasia. Thus, in infants with apnea, laryngospasm, upper airway obstruction, and a typical development of what appears to be bronchopulmonary dysplasia, gastroesophageal reflux must be considered as a possible cause.
REFERENCES
1. Farrell PM, and Avery ME: Hyaline membrane disease, Am Rev Respir Dis 3:657, 1975. 2. Farrell PM, and Wood RE: Epidemiology of hyaline membrane disease in the United States: Analysisof national mortality statistics, Pediatrics 58:167, 1976. 3. Daily W, Klaus M, and Meyer H: Apnea in premature infants, Pediatrics 43:510, 1969. 4. Kanwinkel J, ,Nearman HS, Fanaroff AA, Katona PG, and Klaus MH: Apnea of prematurity, J PEDIATR 86:588, 1975. 5. Watts JL, Ariagno RL, and Brady JP: Chronic pulmonary disease in neonates after artificial ventilation: Distribution of ventilation in pulmonary interstitial emphysema, Pediatrics 60:273, 1977. 6. Klaus M: Respiratory function and puhnonary disease in the newborn, in Barnett H, editor: Pediatrics, New York, 1972, Appleton-Century-Crofts, pp 1255-61. 7. Johnson DG, Herbst JJ, Oliveros MA. and Stewart DR: Evaluation of gastroesophageal reflux surgery in children, Pediatrics 59:62, 1977. 8. Euler AR and Ament ME: Detection of gastroesophageal reflux by tuttle test, Pediatrics 60:65, 1977. 9. Jolley SG, Johnson DG, Herbst JJ, Pena RA, and Garnier R: An assessment of gastroesophageal reflux in children by extended pH monitoring of the distal esophagus, Surgery 84:16, 1978. 10. Schafl'erAJ and Avery ME: Disease of the newborn, ed 4, Philadelphia, 1977, WB Saunders Company, p 315. 11. Avery GB, and Fletcher AB: Nutrition, in Avery GB: Neonatology pathophysiology and management of the newborn, Philadelphia, 1975, JB Lippincon Company, p 878. 12. Johnson P: Laryngeal induced apnea, in Robinson RR. editor: SIDS 1974, proceeding of the Francis E. Camps international symposium on sudden and unexpected deaths
768
13.
14.
Herbst, Minton, and Book
The Journal of Pediatrics November 1979
in infancy, Canada, 1974, The Canadian Foundation for the Study Of Infant Death, p 878. Harned HS Jr, Myracle J and Ferreiro J: Respiratory suppression and swallowing from introduction of fluids into the laryngeal region of the lamb, Pediatr Res 12:1003, 1978. Behar J, Biancani P and Sheahan DG: Evaluation of
15.
esophageal tests in the diagnosis of reflux esophagitis, Gastroenterology 71:9, 1976. Ariagno RL, Guilleminault C, and Negel LE: Mixed and obstructive sleep apnea in 3-month-old control and near miss for sudden infant death syndrome infants, Pediatr Res 12:519. 1978.
Copyright information The appearance of a code at the bottom of the first page of an original article in this JOURNAL indicates the copyright owner's consent that copies of the article may be made for personal or internal use, or li)r the personal or internal use of specific clien;t~:.SThisconsent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc., P.O. BOx 765, Schenectady, N.Y. 12301, (518) 374-4430, for copyin~ beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying l~or general distribution, for advertising or promotional purposes, for creating new collective works, or for resale.
