Skip Navigation LinksHome > June 2011 - Volume 52 - Issue 6 > Utility of Stool Sample–based Tests for the Diagnosis of Hel...
Journal of Pediatric Gastroenterology & Nutrition:
doi: 10.1097/MPG.0b013e3182077d33
Original Articles: Gastroenterology

Utility of Stool Sample–based Tests for the Diagnosis of Helicobacter pylori Infection in Children

Leal, Yelda A*; Cedillo-Rivera, Roberto*; Simón, J Abraham; Velázquez, Juan R; Flores, Laura L§; Torres, Javier||

Free Access
Article Outline
Collapse Box

Author Information

*Unidad de Investigación Médica Yucatán (UIMY), Unidad Médica de Alta Especialidad de Mérida, Instituto Mexicano del Seguro Social, Mérida, Yucatan, México

BIOCEM/HRAEPY Mérida Yucatán

Departamento de Inmunogenética y Alergia, Instituto Nacional de Enfermedades Respiratorias, México, City

§Division of Pulmonary and Critical Care Medicine, San Francisco General Hospital and the University of California, San Francisco, USA

||Unidad de Investigación Médica en Enfermedades Infecciosas y Parasitarias (UIMEIP) Hospital de Pediatría Centro Médico Nacional Siglo XXI, Instituto Mexicano del Seguro Social, México City, Mexico.

Received 30 December, 2009

Accepted 18 October, 2010

Address correspondence and reprint requests to Yelda A. Leal, Unidad de Investigación Médica Yucatán, Unidad Médica de Alta Especialidad del Centro Médico Nacional “Ignacio García Téllez” Mérida Yucatán, Instituto Mexicano del Seguro Social, Calle 34 #439 x 41 Colonia Industrial (Ex-terrenos del Fénix, Hospital T1), Mérida 97150, Yucatán, México (e-mail: yelda_leal03@yahoo.com.mx).

The authors report no conflicts of interest.

Collapse Box

Abstract

Objective: Helicobacter pylori antigen or DNA in stool are meant to detect the bacteria; however, in children the colonization of the gastric mucosa by H pylori is usually weak and fecal excretion of antigen or DNA varies considerably, challenging the utility of these tests in this age group. The aim of the present study was to carry out a systematic review and meta-analysis to evaluate the performance of stool H pylori DNA and antigen tests for the diagnosis of infection in children.

Methods: We conducted a systematic review and meta-analysis to assess the accuracy of stool tests for diagnosis of H pylori infection in children. We searched PubMed, EMBASE, and LILACS databases. Selection criteria included participation of at least 30 children and the use of a criterion standard for H pylori diagnosis. In a comprehensive search, we identified 48 studies.

Results: Regarding antigen-detection tests, enzyme-linked immunosorbent assay (ELISA) monoclonal antibodies showed the best performance, with sensitivity and specificity of 97%, positive likelihood ratio (LR+) of 29.9, and negative likelihood ratio (LR−) of 0.03. ELISA polyclonal antibodies had sensitivity of 92%, specificity of 93%, LR+ of 16.2, LR− of 0.09, and high heterogeneity (P < 0.0001). One-step monoclonal antibody tests demonstrated sensitivity of 88%, specificity of 93%, LR+ of 10.6, and LR− of 0.11. For DNA detection, polymerase chain reaction–based test showed sensitivity of 80.8%, specificity of 98%, LR+ of 17.1, and LR− of 0.18.

Conclusions: Detection of H pylori antigen in stools with ELISA monoclonal antibodies is a noninvasive efficient test for diagnosis of infection in children. One-step tests showed low accuracy and more studies are needed to obtain a useful office-based screening test. The available molecular tests are still unreliable.

Both invasive (eg, endoscopy with biopsy for histology, culture, and rapid urease test [RUT]) and noninvasive (eg, urea breath test [13C-UBT], stool antigen test) methods are available to diagnose Helicobacter pylori infection in children (1–3). Ideally, a diagnostic test for H pylori infection in children should be noninvasive, highly sensitive and specific, inexpensive, and easy to perform. Because H pylori and/or its macromolecules such as proteins and DNA are shed in feces, stool-based tests such as culture, antigen, or DNA detection have become accepted techniques for noninvasive diagnosis. The use of stool culture for isolation of H pylori has limitations because the organism is usually already dead or present in a nonviable coccoid form (4,5). This leaves either stool H pylori DNA or antigen-detection tests as more suitable diagnostic tools (6–8).

Several commercial enzyme-linked immunosorbent assay (ELISA) tests for H pylori stool antigen are available. The main differences among these tests are the nature of the detecting antibodies; some kits use a polyclonal anti–H pylori capture antibody, such as Premier Platinum HpSA ELISA (Meridian Diagnostic, Cincinnati, OH), whereas other assays use monoclonal antibodies such as Amplified IDEIA HpStAR (Dako, Glostrup, Denmark) and Femtolab (Martinsried, Germany). Recently, a new, rapid, 1-step test (immunochromatographic format) using monoclonal antibodies has been introduced, (ImmunoCard STAT! HpSA Meridian Diagnostic, and RAPID Hp StAR, Dako). This test is easy, requires only 5 minutes to complete, and can be performed in a doctor's office (9–13).

Tests that detect H pylori DNA are based on polymerase chain reaction (PCR) amplification of target sequences such as ureA, ureC, and 16S rRNA genes (7,14). In vitro studies with reference strains have shown that although coccoid forms are not viable, they contain DNA and RNA material that is readily detected by PCR (5,14,15). However, PCR-based tests depend on the quality and amount of DNA recovered, the target sequences, and the nature of the amplification protocol, such as nested PCR, real-time PCR, and the like. Several population-based studies evaluating the use of PCR-based tests have reported with variable and inconsistent results, with sensitivities ranging from 25% to 100% and specificities ranging from 80% to 100% (16–19).

Given the diversity of stool-based diagnostic tests for H pylori infection, a critical appraisal of their performance is needed. Several meta-analyses on the topic have been published based predominantly on adult populations and populations of adults and children combined (6,20,22). One meta-analysis of the performance of the commercial test HpSA including 43 studies reported sensitivity and specificity values of 92.4% and 91.9%, respectively, using 13C-UBT as the criterion standard (6). Another meta-analysis of HpSA that only included 6 studies enrolling patients with bleeding peptic ulcer reported lower sensitivity (87%) and specificity (70%) (21). Another recent meta-analysis with 22 studies found that the commercial monoclonal anti–H pylori capture antibody (IDEIA HpStAR) achieved high sensitivity (94%) and specificity (97%) (22). We are unaware of a meta-analysis of the accuracy of stool-based tests in which studies included only children as participants. An initial search of the literature identified 64 studies evaluating H pylori antigen or DNA stool tests in children. Therefore, we carried out a systematic review and meta-analysis to evaluate the performance of stool H pylori DNA and antigen tests for the diagnosis of infection in children.

Back to Top | Article Outline

METHODS

We followed standard guidelines and methods for systematic reviews and meta-analysis of diagnosis (23–26).

Back to Top | Article Outline
Search Strategy

We systematically searched PubMed, EMBASE, and LILACS (June 1998–May 2007) to identify studies evaluating the accuracy of stool H pylori DNA or antigen tests for the diagnosis of H pylori infection in children from birth and up to 19 years old. The search terms used included “Helicobacter pylori,” “children,” “antigen detection,” “diagnostic,” “stool samples,” “nucleic acid amplification,” “polymerase chain reaction,” “sensitivity,” “specificity,” “feces,” and “detection.” In addition, we checked references of eligible articles and corresponded with authors when a full-length article was not available directly online.

Back to Top | Article Outline
Study Eligibility

We included articles that met the following predetermined criteria: comparison of H pylori stool tests with a reference standard (including culture, histological examination, or UBT) (27) and evaluation of a minimum of 30 participants. Although our initial literature search had no language restrictions, studies not available in either English or Spanish were excluded. In addition, we excluded reviews, letters to the editor, opinions, and recommendations about the diagnosis of H pylori infection in children.

Back to Top | Article Outline
Outcomes

Our analysis focused on the following measures of diagnostic accuracy: sensitivity (true-positive results), defined as the proportion of tests determined to be correctly identified as H pylori positive by a reference standard (culture, histology, or UBT); specificity (true-negative results), defined as the proportion of tests determined to be negative; positive likelihood ratio (LR+), which measures how many times a positive test is more likely found in infected versus noninfected children; negative likelihood ratio (LR−), which measures how many times a negative result is more likely found in infected versus noninfected children; and diagnostic odds ratio (DOR), defined as the ratio of the odds of a positive test result in a child with infection compared with a child without infection. The DOR combines sensitivity and specificity into a single measure of test accuracy, ranging from 0 to infinity, with higher values indicating better discriminatory test performance or higher accuracy (24,28).

Back to Top | Article Outline
Data Extraction

Two reviewers (J.A.S. and J.V.R.) extracted data from all of the included articles using a predefined database in Microsoft Excel format. A second reviewer (L.L.F.) independently extracted data of a subset (20 articles [32%]) of the included articles. The interrater agreement between the 2 reviewers for sensitivity and specificity was 95%; differences were solved by consensus. The following data were extracted: study design; criterion standard for H pylori diagnosis; type of test: ELISA and immunochromatography assays for H pylori antigen detection and PCR for DNA detection; direction of data collection (prospective or retrospective); number of participants; age; language (English or Spanish); and data sufficient to construct a 2 × 2 table for true-positives, false-positives, true-negatives, and false-negatives. With this information we calculated for each study the following values: sensitivity, specificity, LR+, LR−, and DOR and their corresponding 95% confidence intervals (95% CI).

Back to Top | Article Outline
Assessment of Study Quality

The quality of the studies was validated using reported tools (25,26) for quality assessment of diagnostic studies: was there a comparison of the H pylori stool test with an appropriate reference standard? was the H pylori stool test result performed in a blinded manner? did the whole sample or a randomly selected subset of the sample receive verification using the reference standard? and did the study recruit prospectively consecutive children suspected of having H pylori infection?

Back to Top | Article Outline
Data Synthesis and Meta-analysis

Meta-analysis was done with a fixed-effects model using MetaDiSc Beta-1.4 software (Universidad Complutense, Madrid, Spain) (29). Forest plots were created to display estimates of sensitivity and specificity and to examine heterogeneity (variability between studies). We summarized the joint distribution of true-positive rate (TPR) and true-negative rate (1 false-positive rate [FPR]) with a summary receiver operating characteristic curve (SROC). The SROC curve in studies of diagnostic accuracy represents the relation between TPR and FPR across studies when test performance is evaluated at varying diagnostic thresholds (30). Each study is a separate unit of analysis and contributes an estimate of TPR and FPR. The overall diagnostic performance of a test can be judged by the position and appearance of the SROC, which is fitted by using a regression model. The area under the curve (AUC) represents an overall summary measure of the curve and the test's overall ability to accurately distinguish cases from noncases. Whereas an AUC of 1 represents perfect discriminatory ability, the Q* index represents a summary of test performance in which sensitivity and specificity are equal. A Q* index of 1.0 indicates perfect accuracy (sensitivity and specificity of 100%). Both AUC and Q* range from 0 to 1, with higher values indicating better test performance (24,30).

Back to Top | Article Outline
Selection of Subgroups for the Meta-analysis and Exploration of Heterogeneity

Heterogeneity in meta-analysis refers to the degree of variability among study results (eg, variability in sensitivity estimates). Heterogeneity may result from differences in threshold (eg, cutoff values), disease spectrum, variation in test protocols, and study quality across studies. When significant heterogeneity is present, summary estimates from meta-analyses are hard to interpret. In this meta-analysis we recognized that studies were heterogeneous in several ways, including the type of diagnostic stool test and the different protocols for anti–H pylori antigen detection. Therefore, we defined subgroups for the meta-analysis by dividing studies into antigen-detection tests (includes 6 different commercial tests) and DNA-detection tests (includes 6 different target sequences). We further divided the antigen-detection subgroup according to the type of antibody-capture test into the following: ELISA polyclonal, ELISA monoclonal, or 1-step monoclonal (Fig. 1). To explore reasons for heterogeneity, the polyclonal ELISA subgroup was further divided by storage temperature (−20°C vs −70°C); blinded status (at least single-blinded interpretation vs not reported); type of criterion standard (culture/histology vs histology/RUT vs UBT); cutoff value (fixed by ROC curve vs not fixed); and number of filters used for ELISA optical density reading (single vs dual wavelengths vs not reported). Heterogeneity was assessed using the χ2 test with MetaDiSc Beta-1.4 software (Universidad Complutense) (29).

Figure 1
Figure 1
Image Tools
Back to Top | Article Outline

RESULTS

Study Selection

The literature search identified 405 potentially relevant citations, of which 39 articles (37 English- and 2 Spanish-language articles) were included (Fig. 2) (31,32). Eight (20.5%) of these reported the evaluation of >1 diagnostic test. In these cases, each test comparison was counted as a separate study. Thus, a total of 48 test comparisons (hereafter referred to as studies) were included in the meta-analysis.

Figure 2
Figure 2
Image Tools
Back to Top | Article Outline
Characteristics of Included Articles

The criterion standards reported among the 39 articles varied: 22 (56.4%) used both H pylori culture and histology (9,11–13,31,33–49), 5 (12.8%) reported only histology and either RUT or UBT (32,50–53), 8 (20.5%) used UBT alone (10,54–60), and 4 (10.3%) used UBT combined with ELISA (61–64). Blinding status also varied among the articles. Seventeen (43.6%) articles reported at least single-blinded interpretation of results, whereas 22 (56.4%) articles did not state blinding status. One (2.5%) (64) article that assessed the performance of PCR reported the use of a randomly selected subset of the sample for validation with the reference standard. All of the articles reported studies that used a case-control design and prospective data collection. A total of 5799 children were included in the meta-analysis. Of these, 2323 were H pylori positive (cases) and 3476 were H pylori negative (controls) compared with the respective criterion standard. The median sample size was 107 participants (interquartile range 66–176).

Back to Top | Article Outline
H pylori Antigen Detection
ELISA Polyclonal Antibody

Twenty-nine studies (3853 children) evaluated the accuracy of antigen detection by the ELISA polyclonal anti–H pylori capture test. The majority of studies assessed the performance of Premier Platinum HpSA (Meridian Diagnostic) and 1 study, the performance of HpSA-Equipar (Equipar Diagnostici, Saronno, Italy). All summary measures of test accuracy were high (Table 1). Overall accuracy was also high, with an AUC of 0.98 and a Q* of 0.94 (Fig. 3). However, as seen in the corresponding Forest plots (Fig. 4), sensitivity (67%–100%) and specificity (61%–100%) values were variable, suggesting the presence of significant heterogeneity (discussed below).

Table 1
Table 1
Image Tools
Figure 3
Figure 3
Image Tools
Figure 4
Figure 4
Image Tools
Back to Top | Article Outline
ELISA Monoclonal Antibody

Six studies (771 children) assessed the performance of commercial ELISA monoclonal anti–H pylori capture tests, 4 studies evaluated Amplified IDEIA HpStAR (Dako), and 2 studies evaluated Femtolab. Pooled measures of performance were high (Table 1). Values of AUC (0.99) and Q* (0.96) were also high (Fig. 5A). Although sensitivity estimates were consistently high (93%–98%), the specificity estimates (88%–99%) were variable (Fig. 6). The test of heterogeneity for the DOR was of borderline significance (P = 0.05; Table 1).

Figure 5
Figure 5
Image Tools
Figure 6
Figure 6
Image Tools
Back to Top | Article Outline
One-Step Monoclonal Antibody

Six studies (779 children) assessed the performance of immunochromatography tests: 5 studies with the ImmunoCard STAT! HpSA and 1 with the RAPID Hp StAR. Pooled accuracy estimates showed moderate performance (Table 1). The AUC of 0.97 and Q* of 0.92 indicated high overall accuracy (Fig. 5B). The DOR was 95.5 (95% CI 38–240.1) and the corresponding test for heterogeneity was significant (P = 0.016).

Back to Top | Article Outline
H pylori DNA Detection

The pooled accuracy of nucleic acid detection tests was assessed in 7 studies (396 children). Three of the 7 studies used a single-reaction PCR protocol; of these, 2 studies used immunomagnetic-bead separation (57,61), 1 study used real-time PCR (43), 2 studies used radiolabeled oligonucleotides (64), and the last study used a seminested-PCR test (63). The amplified target genes were as follows: 16sRNA (1), 23sRNA (2), ureA (1), vacA (1), cagA (1), and 1 test based on the DNA sequence of a species-specific protein antigen (Table 1). Different methods for nucleic acid extraction were used: 2 (28.5%) studies used in-house techniques and 5 (71.5%) studies used commercial kits. As seen in Table 1, specificity estimates were high (93%–100%), whereas sensitivity estimates were variable (54%–100%). The overall accuracy of this group was high, with an AUC of 0.98 and a Q* of 0.94 (Fig. 7). The pooled DOR was 138.4 (95% CI 47.8–400.6), and the corresponding test for heterogeneity was not significant (P = 0.63; Table 1).

Figure 7
Figure 7
Image Tools
Back to Top | Article Outline
Exploration of Heterogeneity

We explored 5 possible causes for heterogeneity: storage temperature, blinded status, type of criterion standard, diagnostic cutoff value, and the use of single- versus dual-wavelength filters among the subgroup of studies using ELISA polyclonal tests. As seen in Table 2, significant heterogeneity was present for storage temperature, blinded studies, and diagnostic cutoff variables, whereas for type of criterion standard and wavelength number, there was no significant heterogeneity. The pooled accuracy parameters in the subgroup of studies using ELISA polyclonal tests should therefore be interpreted with caution.

Table 2
Table 2
Image Tools
Back to Top | Article Outline

DISCUSSION

Principal Findings

This systematic review identified 48 studies that addressed the diagnostic accuracy of stool sample tests for H pylori infection in children. The results of the meta-analysis showed that ELISA monoclonal antibody tests (6 studies) have high pooled sensitivity (97%, 95% CI 94–98) and specificity (97%, 95% CI 95–98) with a significant test for heterogeneity (P = 0.05); although ELISA polyclonal antibody tests (29 studies) yielded high pooled sensitivity (92%, 95% CI 90–93) and specificity (93%, 95% CI 91–94), there was considerable variation in results; 1-step monoclonal antibody tests were less accurate than either polyclonal or monoclonal ELISA tests; in addition, heterogeneity among these studies was significant (P = 0.016), and DNA PCR-based tests yielded variable sensitivities (54%–100%) and high specificities (93%–100%). Because our meta-analysis showed considerable evidence of heterogeneity among studies, we recommend that the results be interpreted with caution.

Back to Top | Article Outline
Antigen-detection Tests

Although there were only 6 studies using the ELISA monoclonal test in children, they reported consistently similar high estimates of sensitivity and specificity (97%). Also, the LRs had good performance: LR+ of 29.9 and LR− of 0.03, meaning that children with H pylori infection have a 29.9-fold greater chance of being positive than children without the infection. This is a good ratio to guide the clinician to rule in (ie, confirm) infection. Similarly, an LR− of 0.03 is sufficiently low to confidently exclude H pylori infection when an ELISA monoclonal test result is negative (65,66). Thus, this ELISA test allows an accurate diagnosis in the absence of annoying and potentially harmful invasive procedures in children, and it is a convenient assay to test H pylori infection in the absence of specific symptoms or even in community-based surveys.

Overall, the ELISA monoclonal antibody tests perform better than the ELISA polyclonal antibody tests, probably because of a higher affinity and specificity of the antibodies; still, specificity values for both tests showed considerable heterogeneity. The possible reasons for this heterogeneity include variability in the study settings, reference standards, and differences in age, genetics, and clinical spectrum of the patients studied (67–69). For example, Frenck et al (37) determined that in children younger than 6 years, the specificity of the ELISA monoclonal antibody test was 81%, whereas in children older than 6 years, specificity increased to 100%.

Of the 3 available H pylori–antigen-detection tests, the 1-step version of the monoclonal antibody test showed the lowest accuracy, a low estimated sensitivity (88%), high heterogeneity, and high LR− of 0.11. With these values, the infection cannot be excluded when test results are negative. An important drawback with the 1-step tests is that they rely on individual visual interpretation of the results, and the interobserver variability and equivocal results are of concern, especially when the test is weakly positive (3,13). In its current form, this test is unreliable and its use is not recommended in children.

Back to Top | Article Outline
DNA-detection Tests

Few studies have addressed the use of PCR amplification for the detection of H pylori in stool samples with the majority being in-house PCR protocols. Our analyses demonstrated that the H pylori DNA PCR detection in stool samples has high specificity (98%) and an acceptable LR+ value (17.1), although sensitivity was low (80.8%) and LR− was high (0.18). One potential explanation for the lower sensitivity in this subgroup is that PCR is susceptible to inhibitor compounds commonly found in feces (70–72). Some authors have reported the use of gene-capture methods to yield more pure H pylori DNA to overcome this problem (57,61,73,74). Another explanation is the increased presence of H pylori coccoid forms from which DNA is more difficult to detect by PCR than from rod-shaped cells (14). To improve the accuracy of the DNA detection tests, we recommend performing nested PCR (17,75,76). In addition, PCR requires specialized laboratory facilities and trained personnel, which are usually not available in diagnostic laboratories in developing countries. Still, an important advantage of this test is the potential to genotype virulence genes and to detect infection with antibiotic-resistance strains in children in whom treatment has failed (43,62,64). Recently, a novel commercial real-time PCR kit (ClariRes, Ingenetix, Vienna, Austria) to detect clarithromycin susceptibility of H pylori in stool specimens was launched. We analyzed 1 study with this assay, in which the reported sensitivity was low (63%) (43). Further studies are needed to evaluate this test in children.

Back to Top | Article Outline
Exploration of Heterogeneity

The shape of the SROC curve (Fig. 3) of the ELISA polyclonal antibody tests suggests that variability in the thresholds could partly explain the heterogeneity (24,30). An additional variable behind this heterogeneity was storage conditions; samples stored at −20°C instead of −70°C improved the accuracy, giving a DOR estimate value 6.2 times greater than before. Inadequate or repeated thawing events may also explain the poor results obtained in some studies. Using UBT as a criterion standard increased sensitivity and specificity to nearly 100% (Table 2); the DOR value was 3.7-fold higher than when H pylori culture/histological examination was used as reference and 4.1-fold higher when the criterion standard was histological examination. These results indicate that UBT makes an appropriate noninvasive reference standard in children (27,77). An appropriate cutoff point represents a crucial factor for the accuracy of the test; no uniformity regarding the cutoff point for positive or negative results was observed across studies included in this meta-analysis. Still, some authors fixed the cutoff value by ROC curve; thus, Bonamico et al (34) reported that 0.160 was an adequate fixed cutoff value in their child population. We analyzed the accuracy of the test before and after fixing the cutoff value, and found that the DOR estimate was about 2 times greater with a fixed cutoff value. This may suggest the need to adjust the cutoff value to the specific studied population when standardizing the protocols. The antigenic differences among the H pylori strains present in the stool of children from different populations may also modify the response to the test. Furthermore, our meta-analysis found that results measured with a dual wavelength showed higher accuracy with a DOR estimate 2.4 times greater than a single wavelength (Table 2). Lastly, a low concentration of H pylori in the stool samples found in infants and younger children may be below the sensitivity of stool-based tests (78).

Although we controlled for different sources of heterogeneity in the studies that used polyclonal antibodies, some of the variation persisted. This may be explained by variability among the different lots of antibodies, to antigenic difference among the H pylori strains used for antiserum production, or to immunological cross-reaction with other Helicobacter species such as H heilmannii, H bilis, H pullorum, H hepaticus, or H felis (79,80). In addition, because the excretion of H pylori antigens in feces is not continuous, test results may vary (81,82).

Back to Top | Article Outline
Clinical and Epidemiological Implications

Based on the results of this meta-analysis, clinicians should consider the detection of antigens in feces using ELISA tests, particularly with monoclonal antibodies, to diagnose H pylori infection in children, especially in young infants because it avoids the need for invasive painful procedures, such as the use of needle to draw blood, or the active collaboration of the child, such as for 13C-UBT. In addition, the test is a noninvasive, simple, rapid test that does not require specialized equipment and is possible to perform at any medical facility. Until recently, treatment decisions in children with severe dyspepsia or chronic abdominal pain were based on the current recommendation by the European Task Force for H pylori Infection in Children (2) and the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition (83), which is to screen symptomatic children with endoscopy because this test allows for the differential diagnosis of abdominal pain including esophagitis, peptic ulcer disease, gastritis, and H pylori infection. Thus, the decision to treat children should include first symptoms, endoscopy findings, and then diagnosis for H pylori infection. Until recently, it was recommended that 13C-UBT and invasive histological examination or culture from biopsies be used because neither antigen nor DNA detection in feces appears sufficiently reliable for clinical use.

H pylori antigen detection tests may be useful for monitoring the success of eradication therapy; furthermore, because stool samples are easily obtained and can be stored frozen, the test is also suitable for epidemiological studies. To achieve better performance with these tests, it is important to carefully evaluate temperature, cutoff value, or wavelength of the assay, and one should also consider that some reports suggest that antimicrobial agents, proton pump inhibitors, and bismuth preparations reduce the sensitivity of these tests (84).

Back to Top | Article Outline
Strengths and Weaknesses of the Review

An important strength of our study is its comprehensive search strategy. Screening, study selection, and quality assessment were done independently by 2 reviewers. For some studies, we reduced the problem of missing data by directly contacting the authors. We also explored heterogeneity and potential publication bias in accordance with published guidelines (24,68,69), and we analyzed data within specific subgroups to lessen the effect of heterogeneity.

We acknowledge that the present study has some limitations; first, we were able to include only English- and Spanish-language articles, and this may have introduced selection bias to our results. Second, we did not address the effect of factors such as laboratory infrastructure, expertise with the technology, and patient demographics. Although we used guidelines such as Standards for Reporting of Accuracy (26) to improve the quality of the analysis, our findings should be interpreted in the context of the quality and variability of the included studies. Unfortunately, we were not able to evaluate the performance of H pylori antigen- or DNA-detection tests in feces among different age groups because this information was not available in the majority of the included studies. The latter issue is relevant because diagnostics test results may vary between young children and infants (11,33,36,44,47).

Back to Top | Article Outline

CONCLUSIONS

The results provided by this systematic review and meta-analysis suggest that detection of H pylori antigens in stool samples with monoclonal antibody–based ELISA test is a noninvasive and efficient test for the diagnosis of H pylori infection in children. The performance of the ELISA test with polyclonal antibodies is inadequate for use in clinical diagnosis and additional adjustments are required. The 1-step polyclonal office-based test showed low accuracy and the protocol should be improved substantially before clinical use. Tests developed to detect DNA are not yet reliable for the diagnosis of H pylori infection in children; still, they may be useful to complement studies of virulence genes and resistance to antibiotics.

Back to Top | Article Outline

Acknowledgments

We thank Karen R. Steingart, MD, MPH, of the Francis J. Curry National Tuberculosis Center, University of California-San Francisco, for her helpful comments and critical revision of the manuscript. J.T. is a recipient of an exclusivity scholarship from Fundacion IMSS, Mexico.

Back to Top | Article Outline

REFERENCES

1. Bourke B, Ceponis P, Chiba N, et al. Canadian Helicobacter Study Group Consensus Conference: update on the approach to Helicobacter pylori infection in children and adolescents–an evidence-based evaluation. Can J Gastroenterol 2005; 19:399–408.

2. Drumm B, Day A, Gold B, et al. Helicobacter pylori and peptic ulcer: Working Group Report of the second World Congress of Pediatric Gastroenterology, Hepatology, and Nutrition. J Pediatr Gastroenterol Nutr 2004; 39(suppl 2):S626–S631.

3. Dzierzanowska-Fangrat K, Lehours P, Megraud F, et al. Diagnosis of Helicobacter pylori infection. Helicobacter 2006; 11(suppl 1):6–13.

4. Cellini L, Allocati N, Angelucci D, et al. Coccoid Helicobacter pylori not culturable in vitro reverts in mice. Microbiol Immunol 1994; 38:843–850.

5. Kusters JG, Gerrits MM, Van Strijp JA, et al. Coccoid forms of Helicobacter pylori are the morphologic manifestation of cell death. Infect Immun 1997; 65:3672–3679.

6. Gisbert JP, Pajares JM. Diagnosis of Helicobacter pylori infection by stool antigen determination: a systematic review. Am J Gastroenterol 2001; 96:2829–2838.

7. Kabir S. Detection of Helicobacter pylori DNA in feces and saliva by polymerase chain reaction: a review. Helicobacter 2004; 9:115–123.

8. van Zwet AA, Thijs JC, Kooistra-Smid AMD, et al. Use of PCR with feces for detection of Helicobacter pylori infections in patients. J Clin Microbiol 1994; 32:1346–1348.

9. Antos D, Crone J, Konstantopoulos N, et al. Evaluation of a novel rapid one-step immunochromatographic assay for detection of monoclonal Helicobacter pylori antigen in stool samples from children. J Clin Microbiol 2005; 43:2598–2601.

10. Hauser B, Wybo I, Tshibuabua G, et al. Multiple-step polyclonal versus one-step monoclonal enzyme immunoassay in the detection of Helicobacter pylori antigen in the stools of children. Acta Paediatr 2006; 95:297–301.

11. Kalach N, Nguyen VB, Bergeret M, et al. Usefulness and influence of age of a novel rapid monoclonal enzyme immunoassay stool antigen for the diagnosis of Helicobacter pylori infection in children. Diagn Microbiol Infect Dis 2005; 52:157–160.

12. Nares-Cisneros J, Jaramillo-Rodriguez Y, Martinez-Ordaz VA, et al. Immunochromatographic monoclonal test for detection of Helicobacter pylori antigen in stool is useful in children from high-prevalence developing country. Helicobacter 2007; 12:354–358.

13. Schwarzer A, Lottspeich C, Russmann H, et al. Evaluation of a novel rapid one-step monoclonal chromatographic immunoassay for detection of Helicobacter pylori in stool from children. Eur J Clin Microbiol Infect Dis 2007; 26:475–480.

14. Narikawa S, Kawai S, Aoshima H, et al. Comparison of the nucleic acids of helical and coccoid forms of Helicobacter pylori. Clin Diagn Lab Immunol 1997; 4:285–290.

15. Bode G, Mauch F, Malfertheiner P. The coccoid forms of Helicobacter pylori. Criteria for their viability. Epidemiol Infect 1993; 111:483–490.

16. Clayton CL, Kleanthous H, Coates PJ, et al. Sensitive detection of Helicobacter pylori by using polymerase chain reaction. J Clin Microbiol 1992; 30:192–200.

17. Li C, Ha T, Ferguson DA Jr, et al. A newly developed PCR assay of H. pylori in gastric biopsy, saliva, and feces. Evidence of high prevalence of H. pylori in saliva supports oral transmission. Dig Dis Sci 1996; 41:2142–2149.

18. Makristathis A, Pasching E, Schutze K, et al. Detection of Helicobacter pylori in stool specimens by PCR and antigen enzyme immunoassay. J Clin Microbiol 1998; 36:2772–2774.

19. Shuber AP, Ascano JJ, Boynton KA, et al. Accurate, noninvasive detection of Helicobacter pylori DNA from stool samples: potential usefulness for monitoring treatment. J Clin Microbiol 2002; 40:262–264.

20. Wisniewska M, Nilsson HO, Bak-Romaniszyn L, et al. Detection of specific Helicobacter pylori DNA and antigens in stool samples in dyspeptic patients and healthy subjects. Microbiol Immunol 2002; 46:657–665.

21. Gisbert JP, Abraira V. Accuracy of Helicobacter pylori diagnostic tests in patients with bleeding peptic ulcer: a systematic review and meta-analysis. Am J Gastroenterol 2006; 101:848–863.

22. Gisbert JP, de la MF, Abraira V. Accuracy of monoclonal stool antigen test for the diagnosis of H. pylori infection: a systematic review and meta-analysis. Am J Gastroenterol 2006; 101:1921–1930.

23. Deville WL, Buntinx F, Bouter LM, et al. Conducting systematic reviews of diagnostic studies: didactic guidelines. BMC Med Res Methodol 2002; 2:9.

24. Lijmer JG, Bossuyt PM, Heisterkamp SH. Exploring sources of heterogeneity in systematic reviews of diagnostic tests. Stat Med 2002; 21:1525–1537.

25. Pai M, McCulloch M, Gorman JD, et al. Systematic reviews and meta-analyses: an illustrated, step-by-step guide. Natl Med J India 2004; 17:86–95.

26. Bossuyt PM, Reitsma JB, Bruns DE, et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. The Standards for Reporting of Diagnostic Accuracy Group. Croat Med J 2003; 44:639–650.

27. Elitsur Y, Tolia V, Gilger MA, et al. Urea breath test in children: the United States prospective, multicenter study. Helicobacter 2009; 14:134–140.

28. Glas AS, Lijmer JG, Prins MH, et al. The diagnostic odds ratio: a single indicator of test performance. J Clin Epidemiol 2003; 56:1129–1135.

29. Zamora J, Abraira V, Muriel A, et al. Meta-DiSc: a software for meta-analysis of test accuracy data. BMC Med Res Methodol 2006; 6:31.

30. Walter SD. Properties of the summary receiver operating characteristic (SROC) curve for diagnostic test data. Stat Med 2002; 21:1237–1256.

31. Gonzalez-Cuevas A, Juncosa T, Jene M, et al. Helicobacter pylori infections: antigen detection in stool samples. Enferm Infecc Microbiol Clin 2001; 19:49–52.

32. Gonzalez FC, Serrano HC, Harris PR. Diagnosis of Helicobacter pylori infection in children based on stool antigen test. Rev Med Chil 2007; 135:182–188.

33. Dondi E, Rapa A, Boldorini R, et al. High accuracy of noninvasive tests to diagnose Helicobacter pylori infection in very young children. J Pediatr 2006; 149:817–821.

34. Bonamico M, Strappini PM, Bonci E, et al. Evaluation of stool antigen test, PCR on ORAL samples and serology for the noninvasive detection of Helicobacter pylori infection in children. Helicobacter 2004; 9:69–76.

35. de Carvalho Costa CL, Rocha GA, Rocha AM, et al. Evaluation of [13C]urea breath test and Helicobacter pylori stool antigen test for diagnosis of H. pylori infection in children from a developing country. J Clin Microbiol 2003; 41:3334–3335.

36. Falsafi T, Valizadeh N, Sepehr S, et al. Application of a stool antigen test to evaluate the incidence of Helicobacter pylori infection in children and adolescents from Tehran, Iran. Clin Diagn Lab Immunol 2005; 12:1094–1097.

37. Frenck RW Jr, Fathy HM, Sherif M, et al. Sensitivity and specificity of various tests for the diagnosis of Helicobacter pylori in Egyptian children. Pediatrics 2006; 118:e1195–e1202.

38. Gosciniak G, Przondo-Mordarska A, Iwanczak B, et al. Helicobacter pylori antigens in stool specimens of gastritis children before and after treatment. J Pediatr Gastroenterol Nutr 2003; 36:376–380.

39. Husson MO, Rolland C, Gottrand F, et al. Evaluation of a Helicobacter pylori stool antigen test for the diagnosis and follow-up of infections in children. Eur J Clin Microbiol Infect Dis 2000; 19:787–789.

40. Kato S, Nakayama K, Minoura T, et al. Comparison between the 13C-urea breath test and stool antigen test for the diagnosis of childhood Helicobacter pylori infection. J Gastroenterol 2004; 39:1045–1050.

41. Koletzko S, Konstantopoulos N, Bosman D, et al. Evaluation of a novel monoclonal enzyme immunoassay for detection of Helicobacter pylori antigen in stool from children. Gut 2003; 52:804–806.

42. Konstantopoulos N, Russmann H, Tasch C, et al. Evaluation of the Helicobacter pylori stool antigen test (HpSA) for detection of Helicobacter pylori infection in children. Am J Gastroenterol 2001; 96:677–683.

43. Lottspeich C, Schwarzer A, Panthel K, et al. Evaluation of the novel Helicobacter pylori ClariRes real-time PCR assay for detection and clarithromycin susceptibility testing of H. pylori in stool specimens from symptomatic children. J Clin Microbiol 2007; 45:1718–1722.

44. Megraud F. Comparison of non-invasive tests to detect Helicobacter pylori infection in children and adolescents: results of a multicenter European study. J Pediatr 2005; 146:198–203.

45. Ni YH, Lin JT, Huang SF, et al. Accurate diagnosis of Helicobacter pylori infection by stool antigen test and 6 other currently available tests in children. J Pediatr 2000; 136:823–827.

46. Oderda G, Rapa A, Ronchi B, et al. Detection of Helicobacter pylori in stool specimens by non-invasive antigen enzyme immunoassay in children: multicentre Italian study. BMJ 2000; 320:347–348.

47. Raguza D, Granato CF, Kawakami E. Evaluation of the stool antigen test for Helicobacter pylori in children and adolescents. Dig Dis Sci 2005; 50:453–457.

48. Roggero P, Bonfiglio A, Luzzani S, et al. Helicobacter pylori stool antigen test: a method to confirm eradication in children. J Pediatr 2002; 140:775–777.

49. van Doorn OJ, Bosman DK, van't Hoff BW, et al. Helicobacter pylori Stool Antigen test: a reliable non-invasive test for the diagnosis of Helicobacter pylori infection in children. Eur J Gastroenterol Hepatol 2001; 13:1061–1065.

50. Sabbi T, De AP, Colistro F, et al. Efficacy of noninvasive tests in the diagnosis of Helicobacter pylori infection in pediatric patients. Arch Pediatr Adolesc Med 2005; 159:238–241.

51. Hino B, Eliakim R, Levine A, et al. Comparison of invasive and non-invasive tests diagnosis and monitoring of Helicobacter pylori infection in children. J Pediatr Gastroenterol Nutr 2004; 39:519–523.

52. Elitsur Y, Lawrence Z, Hill I. Stool antigen test for diagnosis of Helicobacter pylori infection in children with symptomatic disease: a prospective study. J Pediatr Gastroenterol Nutr 2004; 39:64–67.

53. Gulcan EM, Varol A, Kutlu T, et al. Helicobacter pylori stool antigen test. Indian J Pediatr 2005; 72:675–678.

54. Braden B, Posselt HG, Ahrens P, et al. New immunoassay in stool provides an accurate noninvasive diagnostic method for Helicobacter pylori screening in children. Pediatrics 2000; 106:115–117.

55. Kato S, Ozawa K, Okuda M, et al. Accuracy of the stool antigen test for the diagnosis of childhood Helicobacter pylori infection: a multicenter Japanese study. Am J Gastroenterol 2003; 98:296–300.

56. Kato S, Ozawa K, Okuda M, et al. Multicenter comparison of rapid lateral flow stool antigen immunoassay and stool antigen enzyme immunoassay for the diagnosis of Helicobacter pylori infection in children. Helicobacter 2004; 9:669–673.

57. MacKay WG, Williams CL, McMillan M, et al. Evaluation of protocol using gene capture and PCR for detection of Helicobacter pylori DNA in feces. J Clin Microbiol 2003; 41:4589–4593.

58. Rothenbacher D, Bode G, Brenner H. Diagnosis of Helicobacter pylori infection with a novel stool antigen-based assay in children. Pediatr Infect Dis J 2000; 19:364–366.

59. Shepherd AJ, Williams CL, Doherty CP, et al. Comparison of an enzyme immunoassay for the detection of Helicobacter pylori antigens in the faeces with the urea breath test. Arch Dis Child 2000; 83:268–270.

60. Shaikh S, Khaled MA, Islam A, et al. Evaluation of stool antigen test for Helicobacter pylori infection in asymptomatic children from a developing country using 13C-urea breath test as a standard. J Pediatr Gastroenterol Nutr 2005; 40:552–554.

61. Casswall TH, Nilsson HO, Bergstrom M, et al. Evaluation of serology, 13C-urea breath test, and polymerase chain reaction of stool samples to detect Helicobacter pylori in Bangladeshi children. J Pediatr Gastroenterol Nutr 1999; 28:31–36.

62. Booka M, Okuda M, Shin K, et al. Polymerase chain reaction–restriction fragment length polymorphism analysis of clarithromycin-resistant Helicobacter pylori infection in children using stool sample. Helicobacter 2005; 10:205–213.

63. Makristathis A, Barousch W, Pasching E, et al. Two enzyme immunoassays and PCR for detection of Helicobacter pylori in stool specimens from pediatric patients before and after eradication therapy. J Clin Microbiol 2000; 38:3710–3714.

64. Sicinschi LA, Correa P, Bravo LE, et al. Detection and typing of Helicobacter pylori cagA/vacA genes by radioactive, one-step polymerase chain reaction in stool samples from children. J Microbiol Methods 2003; 52:197–207.

65. Pai M, McCulloch M, Enanoria W, et al. Systematic reviews of diagnostic test evaluations: what's behind the scenes? ACP J Club 2004; 141:A11–A13.

66. Deeks JJ, Altman DG. Diagnostic tests 4: likelihood ratios. BMJ 2004; 329:168–169.

67. Menzies D, Pai M, Comstock G. Meta-analysis: new tests for the diagnosis of latent tuberculosis infection: areas of uncertainty and recommendations for research. Ann Intern Med 2007; 146:340–354.

68. Peeling RW, Smith PG, Bossuyt PM. A guide for diagnostic evaluations. Nat Rev Microbiol 2006; 4(12 suppl):S2–S6.

69. Rutjes AW, Reitsma JB, Di NM, et al. Evidence of bias and variation in diagnostic accuracy studies. CMAJ 2006; 174:469–476.

70. Demeke T, Adams RP. The effects of plant polysaccharides and buffer additives on PCR. Biotechniques 1992; 12:332–334.

71. Monteiro L, Gras N, Vidal R, et al. Detection of Helicobacter pylori DNA in human feces by PCR: DNA stability and removal of inhibitors. J Microbiol Methods 2001; 45:89–94.

72. Moreira D. Efficient removal of PCR inhibitors using agarose-embedded DNA preparations. Nucleic Acids Res 1998; 26:3309–3310.

73. Enroth H, Engstrand L. Immunomagnetic separation and PCR for detection of Helicobacter pylori in water and stool specimens. J Clin Microbiol 1995; 33:2162–2165.

74. Monteiro L, Gras N, Megraud F. Magnetic immuno-PCR assay with inhibitor removal for direct detection of Helicobacter pylori in human feces. J Clin Microbiol 2001; 39:3778–3780.

75. Toouli CD, Turner DR, Grist SA, et al. The effect of cycle number and target size on polymerase chain reaction amplification of polymorphic repetitive sequences. Anal Biochem 2000; 280:324–326.

76. Flores LL, Pai M, Colford JM, et al. In-house nucleic acid amplification tests for the detection of Mycobacterium tuberculosis in sputum specimens: meta-analysis and meta-regression. BMC Microbiol 2005; 5:55.

77. Bazzoli F, Cecchini L, Corvaglia L, et al. Validation of the 13C-urea breath test for the diagnosis of Helicobacter pylori infection in children: a multicenter study. Am J Gastroenterol 2000; 95:646–650.

78. Torres J, Perez-Perez G, Goodman KJ, et al. A comprehensive review of the natural history of Helicobacter pylori infection in children. Arch Med Res 2000; 31:431–469.

79. Kornilovs'ka I, Nilsson I, Utt M, et al. Immunogenic proteins of Helicobacter pullorum, Helicobacter bilis and Helicobacter hepaticus identified by two-dimensional gel electrophoresis and immunoblotting. Proteomics 2002; 2:775–783.

80. Pisani P, Whary MT, Nilsson I, et al. Cross-reactivity in immune responses to Helicobacter bilis and Helicobacter pylori in a Thai population at high risk for cholangiocarcinoma. Clin Vaccine Immunol 2008; 15:1363–1368.

81. Leal-Herrera Y, Torres J, Monath TP, et al. High rates of recurrence and of transient reinfections of Helicobacter pylori in a population with high prevalence of infection. Am J Gastroenterol 2003; 98:2395–2402.

82. Perri F, Pastore M, Clemente R, et al. Helicobacter pylori infection may undergo spontaneous eradication in children: a 2-year follow-up study. J Pediatr Gastroenterol Nutr 1998; 27:181–183.

83. Czinn S, Gold B, Dickinson C, et al. Research agenda for Pediatric Gastroenterology, Hepatology and Nutrition: Acid-Peptic Diseases. Report of the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition for the Children's Digestive Health and Nutrition Foundation. J Pediatr Gastroenterol Nutr 2002; 35(suppl 3):S250–S253.

84. Manes G, Balzano A, Iaquinto G, et al. Accuracy of stool antigen test in posteradication assessment of Helicobacter pylori infection. Dig Dis Sci 2001; 46:2440–2444.

Cited By:

This article has been cited 2 time(s).

Advances in Nutrition
Targeted Provision of Oral Iron: The Evolution of a Practical Screening Option
Crowley, CR; Solomons, NW; Schumann, K
Advances in Nutrition, 3(4): 560-569.
10.3945/an.111.001149
CrossRef
Journal of Cancer Research and Clinical Oncology
High circulating VEGF level predicts poor overall survival in lung cancer
Hu, PP; Liu, WD; Wang, LG; Yang, MM; Du, JJ
Journal of Cancer Research and Clinical Oncology, 139(7): 1157-1167.
10.1007/s00432-013-1425-1
CrossRef
Back to Top | Article Outline
Keywords:

diagnosis; Helicobacter pylori; meta-analysis; systematic review

Copyright 2011 by ESPGHAN and NASPGHAN

Login

Article Tools

Images

Share

Search for Similar Articles
You may search for similar articles that contain these same keywords or you may modify the keyword list to augment your search.

Connect With Us

 

 

Twitter

twitter.com/JPGNonline

 

Visit JPGN.org on your smartphone. Scan this code (QR reader app required) with your phone and be taken directly to the site.