Objective. The noninvasive13C-urea breath test (UBT) is a reliable diagnostic method for detection of Helicobacter pylori infection in children, and it avoids invasive gastrointestinal endoscopy. In this study, we compared a noninvasive, newly developed fecal H pylori antigen test with the UBT.
Methodology. One hundred sixty-two children (76 girls and 86 boys) were tested for H pylori infection using the UBT and a new antigen test in stool samples. The H pylori stool test is based on a sandwich enzyme immunoassay with antigen detection.
Results. Twenty-four of the children (14.8%) with dyspepsia tested positive for H pylori according to the breath test results. In 22 of the 24 patients, H pyloriantigen could be detected in the stool (sensitivity: 91.6%). Of 138 patients with negative UBT results, 136 were H pylori-negative in the stool test (specificity: 98.6%).
Conclusions. The new, noninvasive, low-cost H pylori antigen test in stool can replace the UBT for detection of H pylori infection in children with comparable reliability and accuracy.
- UBT =
- 13C-urea breath test •
- HpSA =
- Helicobacter pylori antigen test in stool specimen
During the last 15 years, the role of Helicobacter pylori infection in the pathogenesis of gastritis and peptic ulcer disease in adults1 ,2 and children3 has been elucidated. Therapy of H pylori infection heals peptic ulcer disease.2 ,4 Furthermore, H pylori infection is involved in the pathogenesis of gastric adenocarcinoma and lymphoma.5 ,6
With increased insight into the pathogenicity of H pylori, the demand for a noninvasive, reliable diagnostic test for H pylori infection has emerged. In pediatrics, a noninvasive and practical diagnostic tool for detection of H pyloriinfection is even more desirable, because upper gastrointestinal endoscopies in young children are usually performed in intubation anesthesia or conscious sedation. The disadvantages associated with the endoscopic approach are primarily the invasiveness, the risk of the anesthesia, and discomfort, and can be frightening to the patients and parents. Although diagnostic endoscopy is essential for the primary diagnosis in adult patients with epigastric pain and discomfort (ie, classification of H pylori-dependent disease and detection of H pylori-independent abnormalities), it is not mandatory in children with dyspepsia because malignant gastric diseases are highly unlikely.
Up to the present, the 13C-urea breath test (UBT) is the favorite diagnostic tool in children for the diagnosis ofH pylori infection because it avoids upper gastrointestinal endoscopy. The UBT is based on the stable isotope technique and combines the advantages of noninvasiveness, practicality, excellent sensitivity and specificity, and reproducibility.9–11 The accuracy of the noninvasive UBT in diagnosing H pyloriinfection has also been evaluated in children in reference to histology and culture.12–17 Additionally, the breath test allows a semiquantitative classification of the density of the bacterial colonization on the gastric mucosa.1 ,18
Because many studies support the hypothesis of a fecal–oral route of infection, and because H pylori has been detected in the stool, interest has focused on the diagnostic detection of H pylori antigens in stool samples.
A newly developed H pylori antigen test in stool specimen (HpSA) detects bacterial material in feces.19–22Therefore, it could turn out to be an appropriate noninvasive diagnostic tool for H pylori infection even in children. In this prospective study, we compared the HpSA with the UBT in children with dyspeptic symptoms.
One hundred sixty-two children (76 girls and 86 boys; mean age ± standard deviation: 8.5 ± 3.9 years; age range: 8 months to 15 years) complaining about abdominal pain were screened forH pylori infection using both the UBT and the HpSA.
The study was performed according to the Declaration of Helsinki, and all parents gave informed consent for the participation of their child in the study.
For the UBT,11–13 the patients ingested 75 mg13C-urea (99% atom percent excess) dissolved in 200 mL orange juice (apple juice in children <2 years). Breath samples at baseline and 30 minutes were measured by isotope ratio mass spectrometry (Tracermass, Europa Scientific, Crewe, UK). A δ over baseline value >5 δ‰ indicates H pyloriinfection.
Stool samples were collected on the day of the breath test or 1 day thereafter and were frozen at −20°C until further analysis. The HpSA is based on a sandwich enzyme immunoassay with antigen detection. The HpSA (Meridian Diagnostics, Cincinati, OH) uses polyclonal rabbit anti-H pylori antibodies that are adsorbed to microwells. Stool specimens can be stored for up to 72 hours at 2°C to 8°C in a refrigerator. If testing cannot be performed within this time frame, specimens should be frozen at −20° to −80°C. After thawing to room temperature, small stool particles from the thoroughly mixed stool with diameters of 5 to 6 mm (or 100 μL of liquid stool) are diluted and vortexed in 200 μL of sample diluent (pH 7.2; 10 mmol of phosphate with .02% thimerosal). Fifty-μL aliquots thereof are transferred into the appropriate antibody-coated microwells. The diluted stool specimens are incubated in the microwells together with an enzyme conjugate at room temperature for 1 hour. The enzyme conjugate consists of a rabbit polyclonal antibody specific for H pylori conjugated to horseradish peroxidase in a 50-mmol TRIS-buffered solution with .02% thiomerosal (pH 7.8). Several wash steps (pH 6.8; 180 mmol of phosphate-buffered solution with .2% thimerosal) are performed to remove unbound material. The color reaction starts during the 10-minute incubation with urea peroxide and 3,3′,5,5′-tetramethylbenzidine as substrate in 100-mM citrate–acetate-buffered solution (pH 5.0). The reaction is stopped by adding 2 N sulfuric acid (pH ≤ 1.0). The yellow color intensity was spectrophotometrically read at a wavelength of 450 nm. An optical density OD450 < .140 indicates the absence of H pylori infection. An optical density from .140 to .159 is equivocal and should be repeated according to the manufacturer's instructions. Values ≥.16 indicate the presence of H pyloriantigens.
Twenty-four of the 162 children (14.8%) who were screened forH pylori infection for the first time were H pylori-positive, according to a positive UBT. In 22 of the 24 patients, H pylori antigen could be detected in the stool (sensitivity: 91.6%). One hundred thirty-six of 138 patients with negative UBT results were H pylori-negative in the stool test (specificity: 98.6%). We observed 2 false-negative (13.16 δ‰ vs 9.089; 18.03 δ‰ vs 9.071) and 2 false-positive (4.97 δ‰ vs 9.208; .59 δ‰ vs 9.161) stool test results.
Among all stool analyses, 1 equivocal result (.157) was found that became the second false-positive stool test result after repeating the immunoassay in the stool sample (.161).
One hundred six children were >6 years old. In this age group, 18 children (17.0%) were infected with H pylori according to positive UBT results. The H pylori prevalence in the 56 younger children (≤6 years old) was lower (10.7%). The test qualities of the HpSA in the different age groups are shown in Table 1.
Fifteen of the 24 children (9 girls and 6 boys; age range: 5–13 years) with H pylori infection according to a positive UBT result were given antibiotic therapy to treat dyspeptic symptoms. They were followed up after treatment with omeprazol (1 mg/kg body weight twice daily), clarithromycin (10 mg/kg body weight twice daily), and amoxicillin (25 mg/kg body weight twice daily) for 7 days. Four weeks after the completion of antibiotic therapy, the UBT and the HpSA were reassessed. Five of the 15 children remained H pylori-positive with the UBT, while the stool test revealed only 4 positive results (1 false-negative: 12.87 δ‰ vs .092).
H pylori infection is a common finding in adult and pediatric patients and might be the cause of several gastric diseases. In adults, the endoscopic examination with biopsies from antrum and corpus is the diagnostic method of choice for the primary diagnosis in dyspeptic patients. The UBT is a favorite diagnostic tool to confirm therapeutic success.
In pediatric patients, however, invasive procedures have major disadvantages (risk of anesthesia, discomfort, and frightening to the patients and their parents). Therefore, a noninvasive, practical, and sensitive diagnostic test for the detection of H pyloriinfection is desirable.
A disadvantage of the also noninvasive UBT is the need of expensive analytical equipment, ie, an isotope ratio mass spectrometer, or (relatively) less expensive nondispersive infrared spectrometer is required.12 ,23 This drawback of limited availability, however, has been overcome by analytical centers, organizing a mailing service for test kits. Thus, the UBT has been established in routine diagnostic as a reliable13–17 noninvasive screening method with moderate price. However, its clinical acceptance is more spread in European countries than in the United States.
Serologic methods are unreliable in young children and have been disappointing with respect to the diagnosis of acute infection and control of therapy success after H pylori treatment. The immunologic antibody reaction remains positive for months after successful eradication therapy or spontaneous elimination of the germ. A decline in antibody titers 3 months after eradication therapy may serve as a parameter for successful H pyloritherapy.24 But this is unpractical because both serum samples (before and after therapy) have to be analyzed in 1 batch. Therefore, up to the present, serologic methods are not useful to prove or disprove success of eradication therapy.
The newly developed, noninvasive, enzyme immunoassay HpSA is not time-consuming (the analysis requires ∼90 minutes) and is cheaper than the UBT. The analytical technique of the immunoassay in stool samples can be performed easily in any laboratory. Feces can be obtained easily, even in newborn children. Spot samples of the stool are sufficient; homogenization of the stool is not required.
Because the risk to overlook malignant gastric diseases without performing endoscopies can be neglected in pediatric patients, a screen-and-treat strategy can be recommended in dyspeptic children. Therefore, the fecal antigen test might help to reduce endoscopies (and costs) in children.
In our study, the relatively low eradication rate (66.7%) after treatment of H pylori infection with omeprazole, clarithromycin, and amoxicillin for 7 days might be explained by compliance problems and underlying resistance toward clarithromycin.
According to the findings of our study, the noninvasive HpSA can replace the UBT, thus enabling a low-cost and patient-friendly screening method for H pylori infection. The new HpSA is a highly sensitive and specific, noninvasive diagnostic tool for the qualitative detection of H pylori infection in children.
- Received November 29, 1999.
- Accepted February 23, 2000.
Reprint requests to (B.B.) Medical Department II, University Hospital, Frankfurt/Main, Theodor Stern Kai-7, D-60590, Frankfurt, Germany. E-mail:
- Cullen DJ,
- Collins BJ,
- Christiansen KJ,
- et al.
- Graham DY,
- Klein PD,
- Evans DJ Jr,
- et al.
- Vandenplan Y,
- Blecker U,
- Devreker T,
- et al.
- Makristhasis A,
- Pasching E,
- Schutze K,
- Wimmer M,
- Rotter ML,
- Hirschl AM
- Braden B,
- Teuber G,
- Dietrich CF,
- Caspary WF,
- Lembcke B
- Copyright © 2000 American Academy of Pediatrics