The role of hiatal hernia (HH) as a cause of gastroesophageal reflux (GER) in adult patients has been recently reexamined (1,2). Hiatal hernia contributes to GER and esophagitis by delaying acid clearance from the esophagus (1,2). Gastroesophageal reflux is common in children and associated in some patients with HH. Few studies had been conducted concerning the association of GER and HH in the pediatric population (3). It is unclear whether the presence of HH is a prognostic factor for failure of medical treatment. In 1984, Carre (4) emphasized that the outcome of GER differs in pediatric patients with HH from those without HH. But later, some investigators dismissed the role of HH in pediatric GER disease (5,6). Controversy exists as to whether the presence of HH with GER is an indication for surgical treatment (7–10).
The aims of the present study were 1) to compare 24-hour pH-monitoring studies in children who had GER and HH with children who had GER without HH, and 2) to determine the outcome for children who had GER and HH.
PATIENTS AND METHODS
Between 1989 and 1999, a total of 718 children were diagnosed and treated for GER at our facility. Of these, 45 children (6.3%) had associated HH and became our study group. Based on their neurologic status, the children were divided into two groups. Group A included 35 patients with normal neurologic (NN) development, and group B included 10 children with various severe neurologic disorders (ND), including Canavan disease, metachromatic leukodystrophy, and cerebral palsy. Twenty-seven children with GER and HH underwent esophageal pH monitoring. They were compared with matched controls with GER but without HH who also underwent 24-hour pH monitoring. The groups were matched for age, sex, and neurologic status. Data regarding presenting symptoms, medical history, and treatment modalities were collected from the hospital records.
All the study patients underwent upper gastrointestinal series. The examination was performed in the supine position with a dilute barium solution (E-Z-HD, EZEM™ Inc. Westbury, NY, USA). Each study evaluated the swallowing process, the esophagus, and the stomach, and ended by demonstrating the location of the duodenum and the ligament of Treitz. To demonstrate the presence of HH and/or reflux, the esophagus and the stomach were examined while the patient was rotated to the right and back in supine position. If reflux was not seen, the provocation test of drinking water was performed (siphon test). Hiatal hernia was diagnosed as an upward displacement of the lower esophageal sphincter or identification of more than three gastric mucosal folds above the diaphragm.
The 24-hour pH monitoring was performed as follows: the pH was measured and recorded using a pH recorder (model Digitrapper MK 3, Synectics Medical AB, Stockholm, Sweden). The pH probe was placed at the distal third of the esophagus and verified by chest radiography to be 2 to 3 cm above the diaphragm. The probe was a 2.1-mm OD flexible G91–9011 Monocrystant antimony pH catheter with a reference electrode (Synectics Medical AB, Stockholm, Sweden). Patients received regular feedings. The parents or the nurse recorded the patient's state of wakefulness, mealtimes, and position. The recorded pH data were downloaded into a personal computer. The recordings were analyzed using dedicated software (EsopHogram Software System). The following parameters were compared: number of reflux episodes, number of episodes longer than 5 minutes, fraction of time that pH was lower than 4, esophageal clearance, and reflux index.
Conservative treatment of GER consisted of postural positioning; thickening the formula; and medications including antacids, cisapride, H2-receptor blockers, or more recently proton pump inhibitors. Persistence of vomiting, failure to thrive, apnea episodes, recurrent aspirations, esophagitis, and melena were considered indications of failed medical treatment and indications for surgical intervention. The Nissen or the Thal fundoplications were the surgical antireflux procedures used. Follow-up involved questioning parents or caregivers about general health conditions, developmental abnormalities, rehospitalizations, and investigations, including pH monitoring and treatment.
Analyses were performed using Statistix Statistical Analysis Software version 2.0 (Statistix Analytic Software, La Jolla, CA, USA, 1992). Descriptive statistics for continuous variables such as number of reflux episodes, esophageal clearance, and reflux index are reported as mean ± standard deviation. Frequency counts and cross-tabs were used to describe nominal variables, including the presence or absence of hernia. Pearson correlation coefficients were calculated to describe associations between variables, and the strength of these associations was tested using linear regression analysis (least square method). Spearman rank correlation was used to detect associations in nominal variables and other variables. The t test for independent samples was used to detect differences in reflux variables by hernia status. All tests were considered significant at P < 0.05.
Twenty-four boys and 21 girls were diagnosed with GER and HH. All 45 study patients had sliding HH that were reducible in all children, except for 2 ND patients. The size of the HH was ≤2 cm in 39 patients and 2 cm to 3 cm in 6 patients. Tables 1 and 2 show age at presentation, diagnosis, symptoms, and treatment modalities in the study patients, divided according to neurologic status. All 35 group A (neurologically normal) patients underwent a trial of conservative treatment for GER. Eight group B (neurologic disorder) patients underwent primary surgical treatment because the severity of their reflux and swallowing disorders required permanent gastrostomy feeding. Two group B patients were treated medically because parents refused surgery.
Failure of conservative treatment, manifested by apneic episodes, failure to thrive, or severe esophagitis documented by esophagoscopy (1 patient), was the indication for surgery in all patients treated for GER during the study period. Of the 673 nonstudy patients (i.e., without HH) only 34 (5%) underwent surgery. In contrast, 9 of 35 (25.7%) NN patients in study group A underwent surgery because of failed medical management. Hiatal hernia was not an indication for surgery for any of these patients. The duration of conservative treatment before surgery depended upon the severity of symptoms of GER and ranged between 2 weeks and 12 months (mean, 3.2 months).
Table 3 compares pH-monitoring parameters in 27 patients who had GER and HH with 27 matched controls who had GER and no HH. Each group contained 19 NN patients and 8 ND patients. Esophageal clearance was significantly longer in patients with HH compared with those without HH (P < 0.05). This was consistent when comparing NN study patients with NN control patients (Table 4) and ND study patients with ND control patients (Table 5).
Of 45 study patients, follow-up was available for 30 patients, and ranged from 3 to 54 months (Table 6). Twenty were NN, 9 of whom were treated surgically after medical management failed. All 9 patients underwent surgery without complications and improved clinically. The remaining 11 patients received conservative treatment only. Nine improved clinically, and two are symptomatic and awaiting surgery. All 10 ND patients had follow-up; the 8 patients treated surgically are improved, and the two patients treated medically continue to be clinically symptomatic.
Repeated pH monitoring during the follow-up period was performed in seven NN patients. In four of these, pH was normal and medical treatment was discontinued. In two patients, abnormal results of pH monitoring together with persistent clinical symptoms of GER indicated the need for surgery. In one patient, the pH monitoring was performed after surgery and was normal.
The role of HH in GER has been debated for decades. In the 1950s, Allison (11) was the first to emphasize the association between HH and GER in adults, while Carre (3) demonstrated the importance of HH in the development of childhood GER. The prevalence of HH in children with GER is unknown. According to Foglia (7), HH is uncommon in children with GER. Stewart et al. (12) found the rate of HH in pediatric patients with GER to be 39% and Thomas and Carre (13) found it to be 41%. In our patients with GER, the rate of HH was 6.3%.
In the past few years, the significance of HH has undergone reevaluation. The smooth muscle of the lower esophageal sphincter and the striated muscle of the diaphragmatic crura coordinate to form a sphincter to protect the esophagus against reflux. In patients with HH, alterations of this anatomic relationship could be important factors that contribute to reflux, together with an absence of the flap valve mechanism and the loss of an intraabdominal portion of the esophagus (14).
In our series, which compares patients with and without HH, the results of esophageal pH monitoring showed that tested parameters did not significantly differ, except for esophageal clearance (Table 3). This latter index, which in fact represents the duration of the reflux episodes, was significantly longer in HH patients (P < 0.05) regardless of their neurologic status (Tables 4 and 5). In a study among children who had GER with or without HH, Stewart et al (12) found a longer duration of reflux episodes in children with HH. In fact, the duration of reflux episodes reflects the time that the esophageal mucosa is exposed to the acidic gastric content and, therefore, seems to predict esophagitis. This has been confirmed by endoscopy, both in adults (15) and in pediatric patients (6,12). The mechanism of delayed esophageal clearance in patients with HH has been demonstrated in adult patients (1). Gastric acid is trapped in the hiatal sac in close proximity to the esophagus. Contraction of the crural diaphragm during inspiration and physical maneuvers lead to compartmentalization of the stomach between the lower esophageal sphincter and the diaphragm. Subsequently, acid trapped in the hiatal sac may flow into the esophagus through the lower esophageal sphincter, which is either weak or forced open by negative intrapleural pressure (1,2).
Carre (16) showed that conservative treatment of GER in children with HH before the age of 6 months was correlated with a high success rate. Some authors also report successful results of medical treatment for GER and HH in children (3,10,17). Masliah et al. (18) are less optimistic and found that the presence of HH predicts persistence of reflux symptoms. Stewart et al. (12) also emphasized a possible link between HH and poor prognosis of GER. In long-term follow-up studies among adults who were diagnosed with childhood HH, it was found that, despite clinical improvement, HH persisted in more than 50% of the patients (17,19).
In this series, 9 NN patients (25.7%) underwent surgery because conservative treatment failed (Table 6). At the same time, among the 673 patients diagnosed in our institution as having GER without HH, only 34 (5%) underwent surgery. The 2 NN children with no improvement after medical treatment and the 9 NN patients who received surgery because of failure of medical management represent a significant failure rate of conservative treatment among the NN patients. This failure rate is significantly higher than that of patients with GER without HH. In 2 ND children with HH, conservative treatment of GER also failed. Thus HH in all patients with GER is a marker for potential failure of medical treatment. Repeated pH monitoring may be helpful in deciding to stop medical treatment for patients with clinical improvement of GER. In patients without clinical improvement, repeated pH monitoring may be an important tool in evaluating them for surgery.
In conclusion, we have demonstrated that the presence of HH in children with GER is associated with prolonged exposure of the esophagus to acid. The presence of HH in patients with GER may predict increased failure of medical management. However, because approximately 50% of NN children will respond to a trial of medical treatment, such treatment is warranted in this population. These children need close follow-up with aggressive medical treatment to avoid the development of esophagitis. They also need repeated clinical evaluation and longer follow-up than children who have GER without HH, to facilitate a timely decision about surgical treatment if necessary.
The authors thank Mona Boaz for the statistics analysis of our data.
1. Sloan S, Rademaker AW, Kahrilas PJ. Determinants of gastroesophageal junction incompetence: hiatal hernia
, lower esophageal sphincter, or both? Ann Intern Med 1992; 117: 977–82.
2. Mittal RK, Lange RC, McCallum RW. Identification and mechanism of delayed esophageal acid clearance in subjects with hiatus hernia. Gastroenterology 1987; 92: 130–5.
3. Carre IJ. The natural history of the partial thoracic stomach (hiatus hernia) in children. Arch Dis Child 1959; 34: 344–53.
4. Carre IJ. Clinical significance of gastro-oesophageal reflux. Arch Dis Child 1984; 59: 911–12.
5. Colson DJ, Campbell CA, Wright VA, et al. Predictive value of oesophageal pH variables in children with gastro-oesophageal reflux. Gut 1990; 31: 370–3.
6. Boix-Ochoa J, Lafuenta JM, Gil-Vernet JM. Twenty-four hour esophageal pH monitoring in gastroesophageal reflux
. J Pediatr Surg 1980; 15: 74–8.
7. Foglia RP. Gastroesophageal reflux
. In: Oldham KT, Colombani PM, Foglia RP, eds. Surgery of Infants and Children. Scientific Principles and Practice
. 1st ed. Philadelphia: Lippincott-Raven Publishers; 1997: 1035–47.
8. Chana J, Crabbe DC, Spitz L. Familial hiatus hernia and gastro-oesophageal reflux. Eur J Pediatr Surg 1996; 6: 175–6.
9. Bernhard UA, Shmerling DH. Follow-up examinations of conservatively and surgically treated children with hiatus hernia. Prog Pediatr Surg 1985; 18: 118–31.
10. Guggenbichler JP, Menardi G. Conservative treatment
of gastroesophageal reflux
and hiatus hernia. Prog Pediatr Surg 1985; 18: 78–83.
11. Allison PR. Reflux esophagitis, sliding hiatal hernia
and the anatomy of repair. Surg Gynecol Obstet 1951; 92: 419–31.
12. Stewart RJ, Johnston BT, Boston VE, et al. Role of hiatal hernia
in delaying acid clearance. Arch Dis Child 1993; 68: 662–4.
13. Thomas PS, Carre IJ. Findings on barium swallow in younger siblings of children with hiatal hernia
(partial thoracic stomach). J Pediatr Gastroenterol Nutr 1991; 12: 174–7.
14. Mittal RK, Balaban DH. The esophagogastric junction. N Engl J Med 1997; 336: 24–32.
15. Patti MG, Goldberg HI, Arcerito M, et al. Hiatal hernia
size affects lower esophageal sphincter function, esophageal acid exposure, and the degree of mucosal injury. Am J Surg 1996; 171: 182–6.
16. Carre IJ. Management of gastro-oesophageal reflux. Arch Dis Child 1985; 60: 71–5.
17. Johnston BT, Carre IJ, Thomas PS, et al. Twenty to 40 year follow up of infantile hiatal hernia
. Gut 1995; 36: 809–12.
18. Masliah C, Galmiche JP, Cloarec D, et al. Twenty-four hour oesophageal pH monitoring in infants with suspected gastro-esophageal reflux disease: relationships with symptoms, endoscopy and evolution. Eur J Gastroenterol Hepatol 1990; 2: 137–42.
19. Astley R, Carre IJ, Langmead-Smith R. A 20-year prospective follow-up of childhood hiatal hernia
. Br J Radiol 1977; 50: 400–3.