Surgery for Gastroesophageal Reflux in Children With Normal pH Studies By J.A. Tovar,
J.A. Angulo,
L. Gorostiaga,
and J. Arana
San Sebastik?, Spain 0 Esophageal pH monitoring is recognized as the best diagnostic procedure for gastroesophageal reflux (GER) and operation is seldom recommended in the absence of abnormal pH data. To emphasize that operation should not be ruled out for children who may have false-negative pH studies, we report 14 patients operated on for GER in spite of normal pH-monitoring. The mean age was 54 months (range, 18 to 90). Clinical features included vomiting, dysphagia, respiratory disease, anemia, and torticollis. All had radiologic evidence of GER, and 10 had endoscopic and histological esophagitis. Conventional pH-monitoring values were normal but lower esophageal sphincter pressure and propulsive peristalsis were significantly decreased whereas nonpropulsive contractions were predominant. Operation was recommended after an average of 24 months of unsuccessful medical treatment. Independent postoperative assessment showed that 13 of the 14 patients were relieved of their symptoms and dysphagia persists in one. We suggest that the diagnosis of GER should be accepted on the basis of sound clinical judgement plus more than one abnormal test even when pH results are normal. Operation should not be withheld when clinically indicated. There are several explanations for false-negative pH studies, of which alkaline reflux is probably the most important and warrants further investigation in children. Copyright o 7991 by W.B. Saunders Company INDEX WORDS: gastric.
Gastroesophageal
reflux;
fundoplication,
B
ECAUSE IT IS the only method allowing quantification of both the frequency and duration of acid exposure, pH monitoring of the distal esophagus currently is accepted as the best procedure for the diagnosis of gastroesophageal reflux (GER).le4 This chemical insult is only one of the mechanisms responsible for GER disease and we believe that in some cases reflux cannot be detected by pH studies alone. We have undertaken this study in an attempt to clarify whether the diagnosis of GER should be accepted in the absence of an abnormal pH study to the point of recommending an operation, and whether the results obtained justify such an approach. MATERIALS
AND
METHODS
Since 1982 we have operated on 150 children with GER. Eighteen were operated on despite having extended esophageal pH studies that were within normal limits. Fourteen of these 18 children have been followed-up for more than 2 years after surgery and form the basis of the present study. The remaining 4 were not included because their period of follow-up was shorter, but postoperative results so far have been similar to those reported here. Fourteen children of comparable ages in whom the tentative diagnosis of GER was excluded after complete workup acted as a Journal of Pediatric Surgery, Vol26, No 5 (May), 1991:
pp 541-545
control group for comparison. Their pH monitoring and manometric results were consistent with those found in controls in our former studies.‘,h We felt that invasive procedures were not ethically acceptable in normal children. All patients had a diagnostic workup including detailed clinical history, standard anthropometric measurements, radiographic studies, 24-hour pH studies, manometry, endoscopy. and esophageal biopsies.
Radiographs Upper gastrointestinal barium series were performed by pediatric radiologists experienced in the evaluation of refluxing patients according to generally accepted techniques.
pH Studies Appropriate microelectrodes and pH meters were used. The recordings were stored in a microcomputer and read by means of software developed by ourselves. Our techniques have been previously published.6-“’ The following threshold values were used for defining the presence of abnormal acid exposure: percentage of refluxing time above 4%, more than 16 GER episodes, more than 4 episodes longer than 5 minutes, and episodes longer than 15 minutes. We were able to show in previous studies using receiver operator characteristic (ROC) curves”’ that these values were the best to discriminate between patients with and without GER in our population. In addition, we measured planimetrically the area under pH4 curve as a single-figure variable representative of the frequency, the duration, and the severity of acid exposure.“‘.”
Manometry Tip-occluded, lateral-hole, triple-lumen plastic catheters were used with constant low-flow, high-pressure perfusion, according to previously published protocols.“,‘” We measured lower esophageal sphincter pressure (LESP) and evaluated esophageal body motility with particular attention to nonpropulsive waves (expressed as percentage of the total number of waves) and to propulsive peristalsis (expressed in one single figure [esophageal motor efficiency] obtained by multiplying the frequency of propulsive waves per hour by the mean pressure of those waves in mm Hg). These studies were carried out under basal conditions and after acid challenge.’
From the Universidad de1 Pais Vasco, Hospital N.S. de Aranzazu, San Sebastirin, Spain. Supported in pari by Grant No. 8711605from the Fondo Investigaciones Sanitarias de la Seguridad Social (Madrid, Spain) and the Health Author@ of the Basque Government. Presented at the 37th Annual International Congress of the British Association of Paediattic Surgeons, Glasgow. Scotland, July 25-27, 1990. Address reprint requests to J.A. Tovar, MD, Servicio de Cimgia Pedidtrica, Hospital N.S. de Aranzazu, Apartado 477, 20080, San Sebasticin, Spain. Copyright o 1991 by W B. Saunders Company 0022-3468l91l2605-0009$03.00l0 541
TOVAR ET AL
Endoscopy
Table 1. Symptoms of GER in 14 Children With Normal pH Studies
This was performed with a pediatric fiberendoscope (Olympus GIF P4, Tokyo, Japan). We considered that there was no esophagitis (grade 0) when the distal esophageal mucosa did not appear erythematous. Esophagitis grade I was diagnosed when the mucosa was congestive and bled easily and esophagitis grade II was diagnosed when ulceration, scarring, or retrograde prolapse were observed.
Vomiting
NO.
Percent
11
78.5
Dysphagia/pain
6
42.8
Respiratory tract disease
6
42.8
Anemia
4
28.6
Torticollis (Sandifer’s syndrome)
3
21.4
Brain damage
2
14.3
Growth stunting
2
14.3
Biopsy Several biopsy specimens were taken with appropriate forceps and processed according to previously published protocols.h The same pathologist assigned the lesions to one of four grades of esophagitis. Grade 0 was when the mucosa was normal or had minimal leucocyte infiltration (less than 6 lymphocytes or plasma cells per high power field). Grade I was when infiltration was more conspicuous (less than 12 cells and/or basal layer hyperplasia [more than 3 cell layers or up to % of the total epithelial thickness]). In grade II, infiltration was important (more than 12 cells and/or presence of neutrophils or eosinophils or severe basal hyperplasia). Grade III was diagnosed when there was ulceration. This classification was modified from Leape et al.6.14
Treatment An operation was offered to patients when medical treatment (consisting of upright position, liquid restriction before sleep, and low-volume meals) over long periods (average, 26 months; range, 7 to 64 months) had failed to relieve their symptoms. All families were informed in detail of the inconsistencies of the diagnostic tests and warned about the uncertain results of operation in their particular situation. The operative procedure was Nissen funduplication in all cases.
Assessment All patients have been regularly followed-up in our outpatient clinic and the position and function of the Nissen wrap was radiologically assessed 6 months after operation. After a minimum of 2 years had elapsed, the families were interviewed in another clinic by doctors not associated with the surgical services. Their findings, together with those of their family pediatrician, were used to evaluate the results of treatment.
Statistical Methods Descriptive statistics were obtained for all numerical variables. For comparison between groups, the nonparametric MannWhitney U test was used. Values of P < .05 were accepted as significant throughout. RESULTS
The mean age at diagnosis was 54 months (range, to 90 months). Seven patients were male and 7 were female. Their symptoms are summarized in Table 1. Vomiting, the most common symptom, was severe in 8 children and moderate, accompanying other symptoms, in 3. The 6 children with respiratory tract disease had repeated pneumonia (2 or more episodes per year) and nocturnal cough. Two children with stunted growth were below the third percentile for age-related height. 18
On barium meal examination, all children had massive GER with loss of the angle of His. Hiatal hernia was observed only once. Conventional pH-monitoring variables (percentage of total time below pH4, number of GER episodes, number of episodes lasting more than 5 minutes, and duration of longest episode) were not significantly different between the two groups (Table 2). The area under pH4 curve (AUC) was also significantly different between the two groups, although it was moderately increased in only 4 patients with GER. Manometric results are summarized in Table 2. This test was normal in one case. Ten children had abnormally low LESP (below 10 mm Hg) and 13 had abnormal motility with an increased proportion of non-propulsive waves and impaired peristalsis (low frequency and low pressure of propulsive waves leading to low “basal esophageal motor efficiency”). Acid stimulation of the esophagus did not reverse this esophageal inertia. These data show that the antireTable 2. pH Monitoring and Manometric Results in Control and Refluxing Children Control
Variable
(n = 14)
GER (n = 14)
pH studies Time below pH4 (%)
2.33 + 1.27
2.26 k 1.32
GER episodes (no.)
a.21 r 6.48
14.69 + 16.9
Episodes > 5 min (no.)
0.78+ 0.69
Longest episode (min)
5.91 2 3.6
7.89 + 4.8
9.6 + 6
25.22 k 19.3*
AUC (cm*)
1 + 0.91
Manometric studies LESP (mm Hg)
15.87 + 7.94
10.07 2 6.41”
Non-PW, basal (%)
37.74 + 26.65
66.92 + 27.16* 69.37 k 29.25*
Non-PW, acid (%)
48.68 + 20.78
PW, basal (waves/h)
16.12 ‘- 15.10
6.64 t 7.35*
MPPW, basal (mm Hg)
59.01 k 16.30
49.59 k 17.25
953.57 + 862.57
476.33 k 462.48*
Motor efficiency, basal)t
14.49
a.01+ a.27
PW, acid (waves/h)
16.14?
MPPW, acid (mm Hg)
65.80 + la.31
47.86 + i9.58*
Motor efficiency, acidt
1.064.05 f 1,049.08
464.19 k 376.66*
NOTE. Values given as mean + SD. Abbreviations: AUC, area under pH4 curve; LESP, lower esophageal sphincter pressure;
PW, propulsive waves; acid, hydrochloric
challenge; MPPW, mean pressure PW. lP < .05 v control. tPW (waves/h) x MP (mm Hg). SP < .OOl v control.
acid
SURGERY FOR GER WITH NORMAL
PH STUDIES
flux mechanisms in the present patients were damaged not only at the sphincteric level (first barrier) but at the peristaltic clearing “pump” level (second barrier) as wellI Endoscopic evaluation of the mucosa was normal in 4 patients. Grade I esophagitis was diagnosed in 9 children and grade II in 1. Biopsies were normal in 4 children and showed esophagitis in 10 (8 grade I, 2 grade II). In summary, the diagnosis of GER was made on the basis of clinical history plus three or four abnormal tests in 13 patients. In the remaining patient only two tests were abnormal (radiological studies and manometry). There were no major postoperative complications and both the position and the function of the antireflux wrap-around were adequate 6 months after operation in all patients. Vomiting resolved in all cases, as did respiratory symptoms. All 3 patients with Sandifer torticollis recovered normal neck position after a few months. There had been no height catch-up in the underdeveloped patients. In the early postoperative period, 6 children complained of gas-bloat syndrome, which progressively resolved. Independent assessment through interview with relatives showed that 7 families were “very satisfied” and 6 were “satisfied” with the results of operation. Only one patient was rated as “unsatisfactory” on the basis of persistent dysphagia (this was the patient with only two positive tests, including severely abnormal esophageal motility).
DISCUSSION
Extended pH monitoring has been shown to have high sensitivity and high specificity in adults using healthy volunteers as contro1s’5,16but normal data for healthy children are not available for ethical reasons. Therefore, the specificity and sensitivity for extended pH monitoring at these ages has not been firmly established. In spite of explicit warnings about the limitations of the method in an early report* and repeated reports of false-negative findings of between 10% and 15%,‘7.‘xa widespread belief in the infallibility of this test has resulted in reluctance to accept the diagnosis of GER when pH results are within normal limits. Many years of experience and research using sophisticated tests for GER in children has taught us, like others,” that the diagnosis of GER should be based on sound clinical criteria plus more than one diagnostic test.‘* Having ruled out other causes for the symptoms, the presence of GER can be established
543
by radiography, isotope scanning, or extended pH studies. The motor function of the esophagus can be evaluated by manometry and the integrity of the mucosa can be assessed by endoscopy and biopsy. Not all these test are abnormal in all patients. This present study demonstrates that in this limited group of patients, GER with gastrointestinal, respiratory, and other less frequent manifestations can in fact take place without excessive acid exposure as judged by pH studies. The symptoms observed in these patients did not differ from those common in other children with GER, although torticollis was particularly frequent (21% in this study v 2% in our previous study6). The incidence of esophagitis by endoscopy was similar (71% v 73%) and by biopsy was greater (71% v 59%) than in our series of children with GER and excessive acid exposure demonstrated by pH monitoring.h It is of interest that esophageal motor disorders in these patients were identical to those observed in children with GER and abnormal acid exposure and, in particular, to those who did not respond to medical treatment, namely lowered LESP, impaired peristalsis with predominant nonpropulsive motility, and insensitivity to acid challenge.5 The fact that three children had Sandifer’s syndrome, a rare manifestation of esophageal dysfunction, is further confirmation of dysmotility. We have been very cautious in recommending surgery for these particular children, as demonstrated by the long periods of time between diagnosis and operation, but the results were as good as for other refluxing children. We firmly believe that normal pH studies should not be a contraindication to operation for GER when the clinical history and an adequate number of other tests demonstrate that the diagnosis is firmly established. This study confirms our previous experience, and that of others,“,‘7-‘9 that in a certain number of children with GER, the abnormality is not detected by pH-monitoring studies in spite of an adequate technique, and that the proportion of false-negative results is relatively high as well as difficult to quantify. The reasons for this are as follows. In the first place, the conventional normal limits for pH studies may be inadequate, namely frequency and duration of abnormal acid exposure. As we have previously proposed,““” other techniques for evaluating the pH tracings will identify some of the falsenegative pH studies. For example, 4 of 14 patients had AUC values above normal; these were falsenegative pH studies. Second, GER can remain undetected by pH monitoring in patients in whom gastric acid secretion is
TOVAR ET AL
544
either diminished, as in malnutrition2 or in early infancy,20 or neutralized by feedings. Since we began routinely to perform esophagogastric pH studies we have observed that all infants have long postcibal periods during which gastric acid is buffered by formula in such a way that conventional pH monitoring can be considerably altered. Alkalinization after feeds is a constant finding”0,” without implying that the esophagus will be free of the harmful effects of GER, because, even when the pH is neutral, digestive juice components such as pepsin can damage the mucosa. Third, the liquid refluxed into the esophagus can be alkaline and thus remain undetected by pH monitoring. The potential damaging effect of the duodenogastric reflux phenomenon, in which a failure of pyloric function allows retrograde passage to the stomach of biliary-pancreatic duodenal fluid, has been studied in adults’2~2sand in experimental animals.‘6-“o In this situation the gastric acid can be neutralized, obscuring pH studies, but the risk of esophageal damage from GER may actually be increased. It has been demonstrated in animals that the mucosal lesions are
worst when the refluxing fluid contains bile, pancreatic juice and gastric juice.“-“9 Such evidence is not as convincing in humans, in whom it has been shown that esophagitis is more severe in those patients with very acid GER and slow gastric emptying than in those with alkaline GER:’ although Barrett’s esophagus is more frequent in the latter.2s Regardless of which of these situations occurs in children, it is apparent that they can reduce the diagnostic potential of pH studies. Finally, the conditions of the test themselves (excessive salivation due to the presence of the pH probe) can explain in part why some patients with esophagitis have normal pH results.31 In conclusion, this study confirms our impression that false-negative results in pH monitoring are more frequent than generally believed. When this test is “normal” the diagnosis of GER disease cannot be ruled out and clinical indications for operation should be adhered to when the diagnosis can be firmly established with other methods. This study suggests that it is necessary to develop better methods for evaluating hypoacidity and alkaline GER in children.
REFERENCES 1. Hill JL, Pelligrini CA, Burrington JD, et al: Technique and experience with 24-hour esophageal pH monitoring of the distal esophagus. J Pediatr Surg 12877887, 1977 2. Jolley SG, Johnson DG, Herbst JJ, et al: An assessment of gastroesophageal reflux in children by extended pH-monitoring of the distal esophagus. Surgery 84:16-24.1978 3. Boix Ochoa J, Lafuente JM, Gil Vernet JM: Twenty-fourhour esophageal monitoring in gastroesophageal reflux. J Pediatr Surg 15:74-79,198O 4. Euler AR, Byrne WJ: Twenty-four-hour esophageal intraluminal pH probe testing: A comparative analysis. Gastroenterology 80:957-961,198l 5. Arana J, Tovar JA: Motor efficiency of the refluxing esophagus in basal conditions and after acid challenge. J Pediatr Surg 24:1049-1054, 1989 6. Tovar JA, Blarduni E, Torrado J, et al: Acid exposure and reflux esophagitis. Pediatr Surg Int 3:326-330, 1988 7. Garay J, Arana J, Alzueta MT, et al: Le reflux gastrooesophagien a manifestations respiratoires. Etude manomttrique et pH-metrique. Helv Paediatr Acta 37:221-229,1982 8, Eizaguirre I, Tovar JA, Arana J, et al: Resultats du traitement chirurgical du reflux gastro-oesophagien a manifestations respiratoires. Chir Pediatr 28:20-23, 1987 9. Eizaguirre I, Tovar JA, Emparanza J: Estudio clinico y pH-metric0 de1 Reflujo Gastro-esofagico en nifios con manifestaciones respiratorias. An Esp Pediatr 30:451-456,1989 10. Tovar JA, lzquierdo MA, Eizaguirre I: The area under curve: A single-figure parameter representative of esophageal acid exposure. J Pediatr Surg 26:163-167, 1991 11. Izquierdo MA, Tovar JA, Eizaguirre I: L’exposition acide oesophagienne en un seul chiffre: La surface sous la courbe de pH. Chir Ptdiatr 30:1-5, 1989 12. Arana J, Tovar JA, Garay J: Le Peristaltisme oesophagien et la clearance acide. Une methode d’etude. Chir Pediatr 26:88-91, 1985
13. Arana J, Tovar JA, Garay J: Abnormal preoperative and postoperative esophageal peristalsis in gastroesophageal reflux. J Pediatr Surg 21:711-714,1986 14. Leape the diagnosis
LL, Bhan I, Ramenofski ML: Esophageal biopsy in of reflux esophagitis. J Pediatr Surg 16:379-384, 1981
15. Fuchs ISH, DeMeester TR, AIbertucci M: Specificity and sensitivity of objective diagnosis of gastroesophageal reflux disease. Surgery 102:575-580,1987 16. Schindlbeck NE, Heinrich C, Konig A, et al: Optimal thresholds, sensitivity, and specificity of long-term pH-metry for the detection of gastroesophageal reflux disease. Gastroenterology 9385-90, 1987 17. Vandenplas infants. Evaluation 146:504-507, 1987
Y, Sacre-Smits L: Gastro-esophageal reflux in of treatment by pH monitoring. Eur J Pediatr
18. Meyers WF, Roberts CC. Johnson DG, et al: Value of tests for evaluation of gastroesophageal reflux in children. J Pediatr Surg 20:515-520, 1985 19. Herbst 1981
JJ: Gastroesophageal
reflux. J Pediatr
98859-870,
20. Bocquet A, Bertrand AM, Schirrer J, et al: Inter&t de l’enregistrement simultant des pH oesophagien et gastrique chez le nourrisson. Pediatric 41:17-32, 1986 21. Vandenplas Y, Sacre-Smits L: Continuous 24-hour esophageal pH monitoring in 285 asymptomatic infants O-15 months old. J Pediatr Gastroenterol Nutr 6:220-224, 1989 22. Little AG, Martinez EL, DeMeester TR, et al: Duodenogastric reflux and reflux esophagitis. Surgery 96:447-454,1984 23. Gozzetti G, Pilotti V, Spangaro M, et al: Pathophysiology and natural history of acquired short esophagus. Surgery 102:507514,1987 24. Fei L, DiMartino N, Amato G, et al: Role of combined esophagogastyric twenty-four-hour pH-monitoring in the study of reflux esophagitis and alkaline gastropathy, in Siewert JR, Hol-
SURGERY FOR GER WITH NORMAL
PH STUDIES
scher AH (eds): Diseases of the Esophagus. New York, NY, Springer-Verlag, 1988, pp 839-846 25. Attwood SEA, DeMeester TR, Bremner CG, et al: Alkaline gastroesophageal reflux: Implications in the development of complications in Barrett’s columnar-lined lower esophagus. Surgery 106:764-770,1989 26. Lambert R: Relative importance of biliary and pancreatic secretions in the genesis of esophagitis in rats. Am J Dig Dis 7:1026-1033, 1962 27. Levrat M, Lambert R, Kirshbaum G: Esophagitis produced by reflux of duodenal contents in rat. Am J Dig Dis 7:564-573,1962 28. Henderson RD, Mugashe F, Jeejeebhoy KN, et al: The role
545
of bile and acid in the production of esophagitis and the motor defect of esophagitis. Ann Thorac Surg 14:465-473,1972 29. Safaie-Shirazi S, DenBesten L, Zike WL: Effect of bile salts on the ionic permeability of the esophageal mucosa and their role in the production of esophagitis. Gastroenterology 68:728-733, 1975 30. Evander A, Little AG, Riddell RH, et al: Composition of the refluxed material determines the degree of reflux esophagitis in the dog. Gastroenterology 93:280-286, 1987 31. Schlesinger PK, Donahue PE, Schmid B, et al: Limitations of 24-hour intraesophageal pH monitoring in the hospital setting. Gastroenterology 89:797-804,1985
J.A. Angulo,
L. Gorostiaga,
and J. Arana
San Sebastik?, Spain 0 Esophageal pH monitoring is recognized as the best diagnostic procedure for gastroesophageal reflux (GER) and operation is seldom recommended in the absence of abnormal pH data. To emphasize that operation should not be ruled out for children who may have false-negative pH studies, we report 14 patients operated on for GER in spite of normal pH-monitoring. The mean age was 54 months (range, 18 to 90). Clinical features included vomiting, dysphagia, respiratory disease, anemia, and torticollis. All had radiologic evidence of GER, and 10 had endoscopic and histological esophagitis. Conventional pH-monitoring values were normal but lower esophageal sphincter pressure and propulsive peristalsis were significantly decreased whereas nonpropulsive contractions were predominant. Operation was recommended after an average of 24 months of unsuccessful medical treatment. Independent postoperative assessment showed that 13 of the 14 patients were relieved of their symptoms and dysphagia persists in one. We suggest that the diagnosis of GER should be accepted on the basis of sound clinical judgement plus more than one abnormal test even when pH results are normal. Operation should not be withheld when clinically indicated. There are several explanations for false-negative pH studies, of which alkaline reflux is probably the most important and warrants further investigation in children. Copyright o 7991 by W.B. Saunders Company INDEX WORDS: gastric.
Gastroesophageal
reflux;
fundoplication,
B
ECAUSE IT IS the only method allowing quantification of both the frequency and duration of acid exposure, pH monitoring of the distal esophagus currently is accepted as the best procedure for the diagnosis of gastroesophageal reflux (GER).le4 This chemical insult is only one of the mechanisms responsible for GER disease and we believe that in some cases reflux cannot be detected by pH studies alone. We have undertaken this study in an attempt to clarify whether the diagnosis of GER should be accepted in the absence of an abnormal pH study to the point of recommending an operation, and whether the results obtained justify such an approach. MATERIALS
AND
METHODS
Since 1982 we have operated on 150 children with GER. Eighteen were operated on despite having extended esophageal pH studies that were within normal limits. Fourteen of these 18 children have been followed-up for more than 2 years after surgery and form the basis of the present study. The remaining 4 were not included because their period of follow-up was shorter, but postoperative results so far have been similar to those reported here. Fourteen children of comparable ages in whom the tentative diagnosis of GER was excluded after complete workup acted as a Journal of Pediatric Surgery, Vol26, No 5 (May), 1991:
pp 541-545
control group for comparison. Their pH monitoring and manometric results were consistent with those found in controls in our former studies.‘,h We felt that invasive procedures were not ethically acceptable in normal children. All patients had a diagnostic workup including detailed clinical history, standard anthropometric measurements, radiographic studies, 24-hour pH studies, manometry, endoscopy. and esophageal biopsies.
Radiographs Upper gastrointestinal barium series were performed by pediatric radiologists experienced in the evaluation of refluxing patients according to generally accepted techniques.
pH Studies Appropriate microelectrodes and pH meters were used. The recordings were stored in a microcomputer and read by means of software developed by ourselves. Our techniques have been previously published.6-“’ The following threshold values were used for defining the presence of abnormal acid exposure: percentage of refluxing time above 4%, more than 16 GER episodes, more than 4 episodes longer than 5 minutes, and episodes longer than 15 minutes. We were able to show in previous studies using receiver operator characteristic (ROC) curves”’ that these values were the best to discriminate between patients with and without GER in our population. In addition, we measured planimetrically the area under pH4 curve as a single-figure variable representative of the frequency, the duration, and the severity of acid exposure.“‘.”
Manometry Tip-occluded, lateral-hole, triple-lumen plastic catheters were used with constant low-flow, high-pressure perfusion, according to previously published protocols.“,‘” We measured lower esophageal sphincter pressure (LESP) and evaluated esophageal body motility with particular attention to nonpropulsive waves (expressed as percentage of the total number of waves) and to propulsive peristalsis (expressed in one single figure [esophageal motor efficiency] obtained by multiplying the frequency of propulsive waves per hour by the mean pressure of those waves in mm Hg). These studies were carried out under basal conditions and after acid challenge.’
From the Universidad de1 Pais Vasco, Hospital N.S. de Aranzazu, San Sebastirin, Spain. Supported in pari by Grant No. 8711605from the Fondo Investigaciones Sanitarias de la Seguridad Social (Madrid, Spain) and the Health Author@ of the Basque Government. Presented at the 37th Annual International Congress of the British Association of Paediattic Surgeons, Glasgow. Scotland, July 25-27, 1990. Address reprint requests to J.A. Tovar, MD, Servicio de Cimgia Pedidtrica, Hospital N.S. de Aranzazu, Apartado 477, 20080, San Sebasticin, Spain. Copyright o 1991 by W B. Saunders Company 0022-3468l91l2605-0009$03.00l0 541
TOVAR ET AL
Endoscopy
Table 1. Symptoms of GER in 14 Children With Normal pH Studies
This was performed with a pediatric fiberendoscope (Olympus GIF P4, Tokyo, Japan). We considered that there was no esophagitis (grade 0) when the distal esophageal mucosa did not appear erythematous. Esophagitis grade I was diagnosed when the mucosa was congestive and bled easily and esophagitis grade II was diagnosed when ulceration, scarring, or retrograde prolapse were observed.
Vomiting
NO.
Percent
11
78.5
Dysphagia/pain
6
42.8
Respiratory tract disease
6
42.8
Anemia
4
28.6
Torticollis (Sandifer’s syndrome)
3
21.4
Brain damage
2
14.3
Growth stunting
2
14.3
Biopsy Several biopsy specimens were taken with appropriate forceps and processed according to previously published protocols.h The same pathologist assigned the lesions to one of four grades of esophagitis. Grade 0 was when the mucosa was normal or had minimal leucocyte infiltration (less than 6 lymphocytes or plasma cells per high power field). Grade I was when infiltration was more conspicuous (less than 12 cells and/or basal layer hyperplasia [more than 3 cell layers or up to % of the total epithelial thickness]). In grade II, infiltration was important (more than 12 cells and/or presence of neutrophils or eosinophils or severe basal hyperplasia). Grade III was diagnosed when there was ulceration. This classification was modified from Leape et al.6.14
Treatment An operation was offered to patients when medical treatment (consisting of upright position, liquid restriction before sleep, and low-volume meals) over long periods (average, 26 months; range, 7 to 64 months) had failed to relieve their symptoms. All families were informed in detail of the inconsistencies of the diagnostic tests and warned about the uncertain results of operation in their particular situation. The operative procedure was Nissen funduplication in all cases.
Assessment All patients have been regularly followed-up in our outpatient clinic and the position and function of the Nissen wrap was radiologically assessed 6 months after operation. After a minimum of 2 years had elapsed, the families were interviewed in another clinic by doctors not associated with the surgical services. Their findings, together with those of their family pediatrician, were used to evaluate the results of treatment.
Statistical Methods Descriptive statistics were obtained for all numerical variables. For comparison between groups, the nonparametric MannWhitney U test was used. Values of P < .05 were accepted as significant throughout. RESULTS
The mean age at diagnosis was 54 months (range, to 90 months). Seven patients were male and 7 were female. Their symptoms are summarized in Table 1. Vomiting, the most common symptom, was severe in 8 children and moderate, accompanying other symptoms, in 3. The 6 children with respiratory tract disease had repeated pneumonia (2 or more episodes per year) and nocturnal cough. Two children with stunted growth were below the third percentile for age-related height. 18
On barium meal examination, all children had massive GER with loss of the angle of His. Hiatal hernia was observed only once. Conventional pH-monitoring variables (percentage of total time below pH4, number of GER episodes, number of episodes lasting more than 5 minutes, and duration of longest episode) were not significantly different between the two groups (Table 2). The area under pH4 curve (AUC) was also significantly different between the two groups, although it was moderately increased in only 4 patients with GER. Manometric results are summarized in Table 2. This test was normal in one case. Ten children had abnormally low LESP (below 10 mm Hg) and 13 had abnormal motility with an increased proportion of non-propulsive waves and impaired peristalsis (low frequency and low pressure of propulsive waves leading to low “basal esophageal motor efficiency”). Acid stimulation of the esophagus did not reverse this esophageal inertia. These data show that the antireTable 2. pH Monitoring and Manometric Results in Control and Refluxing Children Control
Variable
(n = 14)
GER (n = 14)
pH studies Time below pH4 (%)
2.33 + 1.27
2.26 k 1.32
GER episodes (no.)
a.21 r 6.48
14.69 + 16.9
Episodes > 5 min (no.)
0.78+ 0.69
Longest episode (min)
5.91 2 3.6
7.89 + 4.8
9.6 + 6
25.22 k 19.3*
AUC (cm*)
1 + 0.91
Manometric studies LESP (mm Hg)
15.87 + 7.94
10.07 2 6.41”
Non-PW, basal (%)
37.74 + 26.65
66.92 + 27.16* 69.37 k 29.25*
Non-PW, acid (%)
48.68 + 20.78
PW, basal (waves/h)
16.12 ‘- 15.10
6.64 t 7.35*
MPPW, basal (mm Hg)
59.01 k 16.30
49.59 k 17.25
953.57 + 862.57
476.33 k 462.48*
Motor efficiency, basal)t
14.49
a.01+ a.27
PW, acid (waves/h)
16.14?
MPPW, acid (mm Hg)
65.80 + la.31
47.86 + i9.58*
Motor efficiency, acidt
1.064.05 f 1,049.08
464.19 k 376.66*
NOTE. Values given as mean + SD. Abbreviations: AUC, area under pH4 curve; LESP, lower esophageal sphincter pressure;
PW, propulsive waves; acid, hydrochloric
challenge; MPPW, mean pressure PW. lP < .05 v control. tPW (waves/h) x MP (mm Hg). SP < .OOl v control.
acid
SURGERY FOR GER WITH NORMAL
PH STUDIES
flux mechanisms in the present patients were damaged not only at the sphincteric level (first barrier) but at the peristaltic clearing “pump” level (second barrier) as wellI Endoscopic evaluation of the mucosa was normal in 4 patients. Grade I esophagitis was diagnosed in 9 children and grade II in 1. Biopsies were normal in 4 children and showed esophagitis in 10 (8 grade I, 2 grade II). In summary, the diagnosis of GER was made on the basis of clinical history plus three or four abnormal tests in 13 patients. In the remaining patient only two tests were abnormal (radiological studies and manometry). There were no major postoperative complications and both the position and the function of the antireflux wrap-around were adequate 6 months after operation in all patients. Vomiting resolved in all cases, as did respiratory symptoms. All 3 patients with Sandifer torticollis recovered normal neck position after a few months. There had been no height catch-up in the underdeveloped patients. In the early postoperative period, 6 children complained of gas-bloat syndrome, which progressively resolved. Independent assessment through interview with relatives showed that 7 families were “very satisfied” and 6 were “satisfied” with the results of operation. Only one patient was rated as “unsatisfactory” on the basis of persistent dysphagia (this was the patient with only two positive tests, including severely abnormal esophageal motility).
DISCUSSION
Extended pH monitoring has been shown to have high sensitivity and high specificity in adults using healthy volunteers as contro1s’5,16but normal data for healthy children are not available for ethical reasons. Therefore, the specificity and sensitivity for extended pH monitoring at these ages has not been firmly established. In spite of explicit warnings about the limitations of the method in an early report* and repeated reports of false-negative findings of between 10% and 15%,‘7.‘xa widespread belief in the infallibility of this test has resulted in reluctance to accept the diagnosis of GER when pH results are within normal limits. Many years of experience and research using sophisticated tests for GER in children has taught us, like others,” that the diagnosis of GER should be based on sound clinical criteria plus more than one diagnostic test.‘* Having ruled out other causes for the symptoms, the presence of GER can be established
543
by radiography, isotope scanning, or extended pH studies. The motor function of the esophagus can be evaluated by manometry and the integrity of the mucosa can be assessed by endoscopy and biopsy. Not all these test are abnormal in all patients. This present study demonstrates that in this limited group of patients, GER with gastrointestinal, respiratory, and other less frequent manifestations can in fact take place without excessive acid exposure as judged by pH studies. The symptoms observed in these patients did not differ from those common in other children with GER, although torticollis was particularly frequent (21% in this study v 2% in our previous study6). The incidence of esophagitis by endoscopy was similar (71% v 73%) and by biopsy was greater (71% v 59%) than in our series of children with GER and excessive acid exposure demonstrated by pH monitoring.h It is of interest that esophageal motor disorders in these patients were identical to those observed in children with GER and abnormal acid exposure and, in particular, to those who did not respond to medical treatment, namely lowered LESP, impaired peristalsis with predominant nonpropulsive motility, and insensitivity to acid challenge.5 The fact that three children had Sandifer’s syndrome, a rare manifestation of esophageal dysfunction, is further confirmation of dysmotility. We have been very cautious in recommending surgery for these particular children, as demonstrated by the long periods of time between diagnosis and operation, but the results were as good as for other refluxing children. We firmly believe that normal pH studies should not be a contraindication to operation for GER when the clinical history and an adequate number of other tests demonstrate that the diagnosis is firmly established. This study confirms our previous experience, and that of others,“,‘7-‘9 that in a certain number of children with GER, the abnormality is not detected by pH-monitoring studies in spite of an adequate technique, and that the proportion of false-negative results is relatively high as well as difficult to quantify. The reasons for this are as follows. In the first place, the conventional normal limits for pH studies may be inadequate, namely frequency and duration of abnormal acid exposure. As we have previously proposed,““” other techniques for evaluating the pH tracings will identify some of the falsenegative pH studies. For example, 4 of 14 patients had AUC values above normal; these were falsenegative pH studies. Second, GER can remain undetected by pH monitoring in patients in whom gastric acid secretion is
TOVAR ET AL
544
either diminished, as in malnutrition2 or in early infancy,20 or neutralized by feedings. Since we began routinely to perform esophagogastric pH studies we have observed that all infants have long postcibal periods during which gastric acid is buffered by formula in such a way that conventional pH monitoring can be considerably altered. Alkalinization after feeds is a constant finding”0,” without implying that the esophagus will be free of the harmful effects of GER, because, even when the pH is neutral, digestive juice components such as pepsin can damage the mucosa. Third, the liquid refluxed into the esophagus can be alkaline and thus remain undetected by pH monitoring. The potential damaging effect of the duodenogastric reflux phenomenon, in which a failure of pyloric function allows retrograde passage to the stomach of biliary-pancreatic duodenal fluid, has been studied in adults’2~2sand in experimental animals.‘6-“o In this situation the gastric acid can be neutralized, obscuring pH studies, but the risk of esophageal damage from GER may actually be increased. It has been demonstrated in animals that the mucosal lesions are
worst when the refluxing fluid contains bile, pancreatic juice and gastric juice.“-“9 Such evidence is not as convincing in humans, in whom it has been shown that esophagitis is more severe in those patients with very acid GER and slow gastric emptying than in those with alkaline GER:’ although Barrett’s esophagus is more frequent in the latter.2s Regardless of which of these situations occurs in children, it is apparent that they can reduce the diagnostic potential of pH studies. Finally, the conditions of the test themselves (excessive salivation due to the presence of the pH probe) can explain in part why some patients with esophagitis have normal pH results.31 In conclusion, this study confirms our impression that false-negative results in pH monitoring are more frequent than generally believed. When this test is “normal” the diagnosis of GER disease cannot be ruled out and clinical indications for operation should be adhered to when the diagnosis can be firmly established with other methods. This study suggests that it is necessary to develop better methods for evaluating hypoacidity and alkaline GER in children.
REFERENCES 1. Hill JL, Pelligrini CA, Burrington JD, et al: Technique and experience with 24-hour esophageal pH monitoring of the distal esophagus. J Pediatr Surg 12877887, 1977 2. Jolley SG, Johnson DG, Herbst JJ, et al: An assessment of gastroesophageal reflux in children by extended pH-monitoring of the distal esophagus. Surgery 84:16-24.1978 3. Boix Ochoa J, Lafuente JM, Gil Vernet JM: Twenty-fourhour esophageal monitoring in gastroesophageal reflux. J Pediatr Surg 15:74-79,198O 4. Euler AR, Byrne WJ: Twenty-four-hour esophageal intraluminal pH probe testing: A comparative analysis. Gastroenterology 80:957-961,198l 5. Arana J, Tovar JA: Motor efficiency of the refluxing esophagus in basal conditions and after acid challenge. J Pediatr Surg 24:1049-1054, 1989 6. Tovar JA, Blarduni E, Torrado J, et al: Acid exposure and reflux esophagitis. Pediatr Surg Int 3:326-330, 1988 7. Garay J, Arana J, Alzueta MT, et al: Le reflux gastrooesophagien a manifestations respiratoires. Etude manomttrique et pH-metrique. Helv Paediatr Acta 37:221-229,1982 8, Eizaguirre I, Tovar JA, Arana J, et al: Resultats du traitement chirurgical du reflux gastro-oesophagien a manifestations respiratoires. Chir Pediatr 28:20-23, 1987 9. Eizaguirre I, Tovar JA, Emparanza J: Estudio clinico y pH-metric0 de1 Reflujo Gastro-esofagico en nifios con manifestaciones respiratorias. An Esp Pediatr 30:451-456,1989 10. Tovar JA, lzquierdo MA, Eizaguirre I: The area under curve: A single-figure parameter representative of esophageal acid exposure. J Pediatr Surg 26:163-167, 1991 11. Izquierdo MA, Tovar JA, Eizaguirre I: L’exposition acide oesophagienne en un seul chiffre: La surface sous la courbe de pH. Chir Ptdiatr 30:1-5, 1989 12. Arana J, Tovar JA, Garay J: Le Peristaltisme oesophagien et la clearance acide. Une methode d’etude. Chir Pediatr 26:88-91, 1985
13. Arana J, Tovar JA, Garay J: Abnormal preoperative and postoperative esophageal peristalsis in gastroesophageal reflux. J Pediatr Surg 21:711-714,1986 14. Leape the diagnosis
LL, Bhan I, Ramenofski ML: Esophageal biopsy in of reflux esophagitis. J Pediatr Surg 16:379-384, 1981
15. Fuchs ISH, DeMeester TR, AIbertucci M: Specificity and sensitivity of objective diagnosis of gastroesophageal reflux disease. Surgery 102:575-580,1987 16. Schindlbeck NE, Heinrich C, Konig A, et al: Optimal thresholds, sensitivity, and specificity of long-term pH-metry for the detection of gastroesophageal reflux disease. Gastroenterology 9385-90, 1987 17. Vandenplas infants. Evaluation 146:504-507, 1987
Y, Sacre-Smits L: Gastro-esophageal reflux in of treatment by pH monitoring. Eur J Pediatr
18. Meyers WF, Roberts CC. Johnson DG, et al: Value of tests for evaluation of gastroesophageal reflux in children. J Pediatr Surg 20:515-520, 1985 19. Herbst 1981
JJ: Gastroesophageal
reflux. J Pediatr
98859-870,
20. Bocquet A, Bertrand AM, Schirrer J, et al: Inter&t de l’enregistrement simultant des pH oesophagien et gastrique chez le nourrisson. Pediatric 41:17-32, 1986 21. Vandenplas Y, Sacre-Smits L: Continuous 24-hour esophageal pH monitoring in 285 asymptomatic infants O-15 months old. J Pediatr Gastroenterol Nutr 6:220-224, 1989 22. Little AG, Martinez EL, DeMeester TR, et al: Duodenogastric reflux and reflux esophagitis. Surgery 96:447-454,1984 23. Gozzetti G, Pilotti V, Spangaro M, et al: Pathophysiology and natural history of acquired short esophagus. Surgery 102:507514,1987 24. Fei L, DiMartino N, Amato G, et al: Role of combined esophagogastyric twenty-four-hour pH-monitoring in the study of reflux esophagitis and alkaline gastropathy, in Siewert JR, Hol-
SURGERY FOR GER WITH NORMAL
PH STUDIES
scher AH (eds): Diseases of the Esophagus. New York, NY, Springer-Verlag, 1988, pp 839-846 25. Attwood SEA, DeMeester TR, Bremner CG, et al: Alkaline gastroesophageal reflux: Implications in the development of complications in Barrett’s columnar-lined lower esophagus. Surgery 106:764-770,1989 26. Lambert R: Relative importance of biliary and pancreatic secretions in the genesis of esophagitis in rats. Am J Dig Dis 7:1026-1033, 1962 27. Levrat M, Lambert R, Kirshbaum G: Esophagitis produced by reflux of duodenal contents in rat. Am J Dig Dis 7:564-573,1962 28. Henderson RD, Mugashe F, Jeejeebhoy KN, et al: The role
545
of bile and acid in the production of esophagitis and the motor defect of esophagitis. Ann Thorac Surg 14:465-473,1972 29. Safaie-Shirazi S, DenBesten L, Zike WL: Effect of bile salts on the ionic permeability of the esophageal mucosa and their role in the production of esophagitis. Gastroenterology 68:728-733, 1975 30. Evander A, Little AG, Riddell RH, et al: Composition of the refluxed material determines the degree of reflux esophagitis in the dog. Gastroenterology 93:280-286, 1987 31. Schlesinger PK, Donahue PE, Schmid B, et al: Limitations of 24-hour intraesophageal pH monitoring in the hospital setting. Gastroenterology 89:797-804,1985
