The Nissen fundoplication procedure is a popular and effective method of surgical management in infants and children with severe gastroesophageal reflux. Dumping syndrome (DS) is a well-recognized complication of this operation. It was first described by Villet et al. (1), and later by others, as a major complication (2,3).
The etiology of the dumping syndrome in these children is poorly understood. Hyperperistalsis with rapid gastric emptying is considered a major factor, but other mechanisms have also been suggested (2,4). Most reported children have had transient symptoms, and few have had the typical complaints described in adults with dumping syndrome following gastric surgery. In our experience, the Nissen fundoplication has been almost uniformly successful (5), but complications after surgery, particularly those requiring further medical or dietary treatment, are common.
This prospective clinical study was specifically designed to evaluate the prevalence of DS following Nissen fundoplication, to determine its relation to gastric emptying, and to advocate a reasonable diagnostic and therapeutic approach.
SUBJECTS AND METHODS
From 1989 to 1993, Nissen fundoplication was performed in 50 consecutive infants and children (29 boys and 21 girls) with severe gastroesophageal reflux. The mean age at the time of operation was 40 months (range, 2 months to 15 years); 22 patients (44%) were below the age of 1 year. Twenty-four patients had major associate disorders—15 (30%) had neurologic impairment, nine suffered from more than one disease, and six had familial dysautonomia.
The preoperative evaluation included nutrition and clinical assessment in all patients, an upper gastrointestinal study in 49 cases, 24-h pH monitoring in 47 cases, and milk scan and technetium scintigraphy for the measurement of gastric emptying in 39 cases. Esophagoscopy was performed in six patients with symptoms of upper GI bleeding and constituted an indication for surgery in two patients.
The fundoplication was performed as originally described by Nissen (6). During the operation, both vagal nerves were identified and protected within the wrap. The hepatic vagal branches were not transected. Pyloroplasty was performed in five patients. A Stamm gastrostomy was concurrently placed in 42 patients. The indication for permanent gastrostomy was severe mental retardation (six patients) and familial dysautonomia (six patients).
In the immediate postoperative period, postprandial glucose levels were examined in all patients with clinical symptoms of DS. In all patients, follow-up of 6 months to 5.5 years after surgery (mean, 26.8 months) included a personal interview with the patients and their parents by an independent dietary specialist. Special emphasis was given to postprandial dumping symptoms and pre- and postprandial complaints including refusal to feed, irritability, nausea, diarrhea, pallor, diaphoresis, weakness, and lethargy. All patients with one or more symptoms were further evaluated by glucose tolerance test, HbA1C level, gastric technetium scintigraphy study, and an upper gastrointestinal study. GTT and gastric emptying study were performed in a control group of 12 nonsymptomatic postfundoplication patients. Patients who showed abnormal blood glucose profiles were diagnosed as suffering from DS and started on specific treatment.
Glucose tolerance test, performed after an overnight fast, consisted of a bolus of 1.75 g/kg glucose given orally. Two patients, older than 18 years, received 100 g glucose. Capillary blood samples were taken for measurement of glucose levels at 0 (fasting), 30, 60, 120, and 180 min after ingestion of the glucose bolus. Hyperglycemia was defined as any glucose level above 166, 151, 130, and 112 mg/dl capillary blood at 30, 60, 120, and 180 min, respectively (7). Hypoglycemia was defined as a capillary blood glucose below 50 mg/dl at any time during the test. HbA1C was measured by the Glyco-Tek affinity column method, utilizing a spectrophotometer operating at 415 nm (Helena Laboratories). The normal range in our laboratory in children up to 18 years is 4.3-7.7%.
Gastric emptying was studied by technetium scintigraphy. The tests were performed at a fixed time following an overnight fast, and any drugs that could have affected stomach motility were stopped 24 h before the examination. The rate of gastric emptying was determined by measuring, at regular time intervals, the radioactivity of the stomach and expressing it as a percentage of the total activity for the entire abdomen. The index employed for gastric emptying was T½—the time in minutes for maximum count to be reduced by 50%. Specific attention was given to the relationship between pre- and postoperative gastric emptying changes in patients with dumping syndrome.
All patients suffering from DS were treated, including parenteral nutrition for those who could not tolerate enteral feeding. Patients under 2 years who could tolerate enteral food were treated with small, frequent meals of lactose-free formula (ProSobe or Isocal, Mead Johnson), with added uncooked cornstarch 3.5-7%. Incremental calorie enrichment was achieved by the addition of 1-3% MCT (mean chain triglycerides). Parents were advised to refrain from giving food and beverages containing glucose. In patients 6 months and older, a chicken-meat formula was given once a day. Patients older than 2 years were advised to wait for 30 min between solid food and beverage intake; to eat small, frequent meals with low glucose content, and to add fats to each meal. In patients whose postprandial symptoms persisted, pectin 5-10 g/day for patients under 12 years and 10-15 g/day for older patients, divided into six doses, was added to the diet.
All patients with DS were reassessed at 3 and 6 months following the diagnosis by personal nutritional and clinical interviews. GTT and HbA1C blood levels were repeated 3 and 6 months following the diagnosis of DS.
Variations in the mean results were shown by standard deviation (SD). Student's t test was used to determine the significance of differences.
In the immediate postoperative course, five patients had severe intolerance to oral feeding, postprandial irritability, diarrhea, diaphoresis, and lethargy. In these patients, postprandial glucose levels were in the range of 310-240 mg/dl (mean, 260; SD ± 61). Based on their striking clinical presentation, and the abnormal postprandial glucose levels, these patients were diagnosed as having immediate severe postoperative DS (SDS). Long-term postoperative clinical assessment of all 50 revealed 18 patients who suffered from one or more symptoms of DS (Table 1). Ten of these, on further evaluation, had an impaired GTT.
The most common postprandial symptoms were feeding difficulties, which appeared in 12 patients. Next most common was nausea in 10; postprandial diarrhea and diaphoresis in nine; weakness, lethargy, and pallor in seven; and vomiting in one patient. The 10 patients with impaired GTT were diagnosed as having delayed latent postoperative dumping syndrome (LDS), and were evaluated further. In all 10, mechanical problems were excluded by postoperative upper gastrointestinal studies. One patient with LDS died from his underlying disease 2 months after diagnosis. In the eight patients with normal GTT, postprandial symptoms were related to underlying medical problems in five cases and postfundoplication complications in three cases, and of these, gas bloat syndrome was found in two, and narrowing of the fundoplication in one.
In all, the symptoms resolved following specific treatment. Two of the five patients who suffered from SDS later developed LDS. The other three patients with SDS had complete resolution of their symptoms 4-6 months after surgery. The preoperative T½ of the control group was within a range of 40-150 min (mean, 84.4; SD ± 33.3). The postoperative T½ was within a range of 40-140 min (mean, 82.8; SD ± 35.7). The variation in gastric emptying between the patient with DS and the nonsymptomatic patients did not show statistical significance.
The result of the GTT in the 10 patients with LDS (Fig. 1) were as follows: at 30 min, glucose levels range was 124-341 mg/dl, with a mean of 224 and SD ± 69.2; at 60 min, the blood glucose levels range was 126-307 mg/dl (mean, 210; SD ± 62.1); at 120 min, 39-400 mg/dl (mean, 99.8; SD ± 106), and at 180 min, 42-400 mg/dl (mean, 100; SD ± 107.7). Seven patients had significant hyperglycemia at 30 min with glucose levels of 210-341 mg/dl (mean, 257; SD ± 53). At 60 min, seven patients had hyperglycemia with a blood glucose level range of 192-310 mg/dl (mean, 243; SD ± 42). Hypoglycemia was found in one patient at 120 min with a glucose level of 39 mg/dl, and in two other patients at 180 min, with glucose levels of 47 mg/dl and 42 mg/dl, respectively.
The results of GTT in the 12 patients from the control group were as follows: glucose levels at 30 min ranged from 69 to 91 mg/dl (mean, 81.3 mg/dl; SD ± 12.43); at 60 min, the range was 88.1-141 mg/dl (mean, 114.8 mg/dl; SD ± 25.06); at 120 min, the range was 79.4-110 mg/dl (mean, 101.9; SD ± 20.26); and at 180 min, glucose levels ranged from 24 to 114 mg/dl (mean, 89.67; SD ± 19.71). Results of GTT in the LDS patients compared to the control group showed statistical significance (p < 0.05) at 30, 60, 120, and 180 min.
In the patients with LDS, high HbA1C levels were found in three patients, ranging from 7.9-9.0% (mean, 8.2; SD ± 0.5). Three patients with mental retardation and two with familial dysautonomia developed DS.
Of the five patients with SDS, two had undergone pyloroplasty. All the patients with SDS and eight with LDS had a postoperative gastrostomy, which remained permanent in three patients from the SDS and two from the LDS groups.
In the 15 patients with dumping syndrome, the range of preoperative T½ was 26-180 min (mean, 81.3; SD ± 41.9), whereas the postoperative T½ ranged from 30 to 120 min (mean, 80.3; SD ± 35.9). In this group, postoperative gastric scan showed faster emptying in six patients, ranging from 35 to 125 min (mean, 65; SD ± 31.6), compared to the preoperative range of 60-180 min (mean, 94.1; SD ± 45.8). There was no specific difference in T½ between patients with SDS and LDS. The variation in gastric emptying did not show statistical significance.
In the immediate postoperative period, one patient with SDS required complete restriction of food, and total parenteral nutrition for 4 weeks, before oral food was resumed.
Following 3 months of dietary treatment, resolution of dumping symptoms was observed in seven of the 14 patients who survived (50%). Six patients (42.8%) showed significant improvement, and one (7.1%) remained unchanged. Following dietary treatment, the GTT returned to normal in two patients. The GTT in two patients improved significantly, with mean glucose levels of 221.5, 170.5, 93, and 59.5 mg/dl, at 30, 60, 120, and 180 min; in five cases, the values did not change.
HbA1C levels remained above normal in two patients (8.2, 10.7%). Following 6 months of dietary treatment, 11 patients (78.5%) enjoyed complete resolution of their symptoms. These patients resumed a normal diet with no recurrence of dumping symptoms. Three (21.4%) showed significant clinical improvement with complete resolution of the postprandial symptoms. GTT returned to normal levels in 11 patients.
Dumping syndrome is a well-recognized complication of gastric surgery in adults, but it is considered to be rare in children. Insufficient attention has been paid to this specific complication of Nissen fundoplication in children (8,9). There have been increasing numbers of reports on dumping syndrome following Nissen fundoplication, mostly as case reports (1,3,4,10,11), indicating that dumping syndrome is more common than previously thought and often goes unrecognized. Our study concerns, 15 patients who underwent Nissen fundoplication and developed dumping syndrome. The prevalence of 30% found in our study is consistent with a report by Hanimann and Hirsig (12), in which seven of 27 patients (25.9%) developed dumping syndrome following Nissen fundoplication.
Dumping syndrome may present with a variety of nonspecific, postprandial symptoms, with as many as two to six symptoms in the same patient (in this study, we found an average of 3.9). The most common symptoms are feeding difficulties, which were seen in 11 of our patients (73.3%). Because the central regulation of feeding is closely linked to blood glucose homeostasis, high blood glucose may have an immediate influence on feeding patterns (4). When feeding difficulties occur following Nissen fundoplication, dumping syndrome should be suspected.
It appears that postfundoplication DS can take two different clinical forms. Since the first pattern, occurring in the immediate postoperative period, is usually associated with severe clinical symptoms, diagnosis is easier. In the second, the clinical presentation is more subtle and the duration longer. In our patients, the interval between operation and diagnosis in the LDS group was a mean of 27.6 months. This lengthy period of time can be attributed to the nonspecificity and variety of LDS symptoms, to the fact that many patients were mainly nonverbal, and to the lack of awareness of LDS on the part of the physicians—although all patients were followed routinely in an out-patient clinic.
A common explanation of dumping syndrome is that rapid gastric emptying results in a high concentration of carbohydrate in the small intestine, which draws plasma into the bowel, causing abdominal fullness, increasing peristalsis, and symptoms of hypovolemia. This mechanism has not yet been tested in large groups of children and infants. However, in adult volunteers, injection of 50% glucose directly into the small intestine produced DS in only one-third of the cases (4). Thus, it seems that other pathogens or mechanisms other than rapid gastric emptying must be involved.
In previously reported children with postfundoplication DS, hyperperistalsis with rapid gastric emptying was found in an upper gastrointestinal study in two cases (3), and a rapid emptying phase was demonstrated by gastric scintigraphy in two other cases with pyloroplasty (11,13). In our patients, pre- and postoperative gastric scintigraphy failed to demonstrate a strict relationship between gastric emptying and dumping syndrome (p > 0.05). These results call into question both the role of rapid gastric emptying in DS and the use of a radio-nuclide gastric emptying study as a diagnostic test in DS.
Whenever DS is suspected, GTT should be considered a simple and reliable diagnostic procedure. The use of HbA1C levels for diagnosis was found to be much less accurate. This latter test is routinely used in diabetic patients to detect and estimate the frequency and severity of recurrent hyperglycemia (14), but it is not very helpful for the diagnosis and follow-up of DS. Only three of 10 of our patients with LDS had abnormally high HbA1C levels that correlated with the GTT and the clinical presentation.
Other factors that may be involved in DS are gastrostomy and pyloroplasty. In our study, a combined fundoplication and gastrostomy was performed in 86.6% of the patients who subsequently developed dumping syndrome. Although gastrostomy probably does not lead directly to DS, it could be a predisposing factor (2).
Disruption of pyloric sphincter function cannot account for the majority of the cases in our study, since pyloroplasty was performed in only two patients who developed dumping syndrome. Furthermore, Buchmiller et al. (15) have advocated the routine use of pyloroplasty with Nissen fundoplication on the grounds that they could demonstrate no correlation between DS and pyloroplasty.
Nor can accidental damage to the vagus nerves while performing the fundoplication account for the majority of cases in this study. During the operation, both vagal nerves were identified in all patients. In addition, DS following vagal nerve damage is not a transient syndrome, yet the majority of our patients showed a marked improvement with time.
Treatment of DS is mandatory to improve the growth of young patients and the quality of life for older ones (16). Dietary regimen has proven itself to be a very effective form of treatment. With 3 months of dietary treatment, there was clinical improvement in 92.8% of the patients; after 6 months, there was clinical improvement in all patients (100%). These results indicate that 3-6 months are required to achieve clinical and GTT improvement. Dietary compliance is essential. In the present study, we found a very good correlation between compliance with the dietary regimen and clinical improvement.
The high prevalence of dumping syndrome following Nissen fundoplication observed in our patients suggests that DS is underdiagnosed in children and infants. Following fundoplication, strict follow-up is mandatory. Even a single clinical symptom, usually feeding difficulties, is an indication for a GTT. This simple, inexpensive diagnostic test is very important. Recognition and awareness of dumping syndrome is necessary for early treatment and a better prognosis.