Evaluation of the Helicobacter pylori stool antigen test (HpSA) for detection of Helicobacter pylori infection in children.

OBJECTIVE

Helicobacter pylori (H. pylori) infection is usually acquired in early childhood. Noninvasive methods for detection of H. pylori infection are required to study its incidence, transmission, and clearance. They should be easy to perform, inexpensive, and have a high diagnostic accuracy, especially in infants and toddlers. Both serology and the 13C-urea breath test (13C-UBT) do not fulfill all these requirements. The aim of this study was to evaluate a new enzyme immunoassay for detection of H. pylori antigen in stool (Premier Platinum HpSA, Meridian Diagnostics, Cincinnati, OH) in a large cohort of children and to compare it to invasive techniques and the 13C-UBT.

METHODS

HpSA was performed in 310 stool samples of 274 children divided into three groups. Group A consisted of 145 children and adolescents (0.5-19.8 yr, 66/145 <6 yr) who underwent upper endoscopy for various gastrointestinal symptoms. H. pylori status was defined by histology, culture, and rapid urease test from biopsies of the antrum and corpus. A 13C-UBT was performed in 133 of 145 children. Group B consisted of 22 patients (5.7-16.1 yr) who were retested with both noninvasive tests 8 wk after anti-H. pylori triple therapy. Group C consisted of 129 healthy infants and toddlers (0.9-3.1 yr) who were tested with the 13C-UBT. Children with discrepant or positive test results were retested after 2 and 12 months. Results of the HpSA were read at 450/620 nm by spectrophotometry. An optical density <0.100 was defined as negative, >0.120 as positive, and values between 0.100 and 0.120 were considered as equivocal.

RESULTS

In Group A, the HpSA gave false-negative results in five of 45 infected children and false-positive results in four of 100 noninfected children, whereas four patients (2.8%) showed equivocal results. In both infected and noninfected children, no relation between the optical density values and age was found. The 13C-UBT was correct in 132 of 133 children tested. In Group B, there was complete concordance between the HpSA and 13C-UBT: 19 children tested negative and three positive. In Group C, concordant results between the two noninvasive methods were found in 124 of 129 (96%) toddlers (122 negative and two positive). Retesting of five children with discrepant results revealed that, on initial testing, the HpSA was incorrect in two (one false-positive, one false-negative), and the 13C-UBT was incorrect in three (always false-positive).

CONCLUSIONS

In symptomatic children, the HpSA revealed a sensitivity of 88.9% (95% CI 77.3-96.3) and a specificity of 94.0% (88.1-97.7) compared to the 13C-UBT, 100% (94.0-100) and 98.9% (94.7-100), respectively. However, in healthy toddlers, the HpSA performed as well as the 13C-UBT with excellent concordance between the two noninvasive tests. There was no age dependency of the stool test results, and changing the cutoff would not have improved accuracy. Thus, the HpSA test seems suitable to monitor the success of anti-H. pylori therapy.