Faecal antigen tests in the confirmation of the effect of Helicobacter eradication therapy

Abstract

Background. The frequent occurrence of Helicobacter pylori infection requires significant health care resources after eradication therapy. Therefore the non‐invasive testing methods are required to alleviate the increased work‐load of health care personnel and to allow an easy control of eradication therapy. Conventionally, the effect of eradication therapy has been confirmed with ¹³C‐urea breath test 4–6 weeks after a completed eradication.

Aim. To assess the applicability of Helicobacter pylori stool antigen tests as alternatives to the breath test in the control of the effect of eradication therapy.

Methods. Fifty patients were diagnosed Helicobacter‐positive by endoscopy and histology as well as by rapid urease test from mucosal specimen. Four weeks after an eradication therapy the patients were subjected to ¹³C‐urea breath test as well as to faecal Helicobacter pylori antigen tests with mono‐ and polyclonal primary antibodies.

Results. The monoclonal and polyclonal stool tests had 94% and 88% sensitivity, and 100% and 97% specificity, respectively, in the detection of Helicobacter pylori infection as compared to the ¹³C‐urea breath test. The non‐invasive test results were completely parallel in patients with various grades of mucosal atrophy or intestinal metaplasia.

Conclusions. Monoclonal faecal Helicobacter pylori antigen test is slightly superior to the polyclonal test regarding the sensitivity in the detection of stool Helicobacter antigens. Due to their sufficient sensitivity and specificity, and to their practicability and cost‐effectiveness, they can be recommended for non‐invasive testing of Helicobacter pylori infection as alternatives to the ¹³C‐urea breath test.

• Diagnostic tests
• eradication therapy
• faecal antigens
• Helicobacter pylori

Introduction

The frequent occurrence of Helicobacter pylori‐positive patients prompts the development of less invasive diagnostic alternatives to conventional endoscopy for the primary detection and control of Helicobacter pylori eradication therapy. Even though in older age groups and in symptomatic patients the primary diagnosis of Helicobacter pylori infection usually requires upper gastrointestinal endoscopy with biopsy, in younger age groups even the primary diagnosis might be made by less invasive diagnostic methods. The diagnosis of Helicobacter pylori infection can be performed by endoscopy and histology with Giemsa staining, rapid urease test, culture, serologic methods or by non‐invasive methods such as ¹³C‐urea breath test and stool tests. No single test can presently be considered as a gold standard, and for clinical studies a rapid urease test, histology and direct demonstration of helicoid micro‐organisms are used as standards 1.

Faecal Helicobacter pylori antigen testing has been reported to serve as a good alternative to more complex testing, such as histologic demonstration of Helicobacter pylori colonization or ¹³C‐urea breath test. Though the acceptable sensitivity and specificity of various antibodies for the detection of stool Helicobacter pylori antigen after eradication therapy have been previously reported, the effect of mucosal morphology on the test result is largely unclear. Therefore, in the present study the monoclonal and polyclonal antibody tests for stool Helicobacter pylori antigen demonstration were compared with the ¹³C‐urea breath test regarding their sensitivity and specificity in the detection of Helicobacter pylori infection in patients who had undergone endoscopy with biopsy and positive rapid urease test as well as histological confirmation of the presence of helicoid micro‐organisms after standard eradication therapy. In addition, the study assessed the influence of mucosal morphology on the results of non‐invasive Helicobacter pylori tests.

Key messages

• Faecal poly‐ or monoclonal testing for Helicobacter antigens is equally sensitive to the ¹³C‐urea breath test in the diagnostic testing for Helicobacter pylori colonization.

Methods

Patients

Fifty patients (female to male ratio 36/14, mean age 58.0 years ) were included in the study after an informed consent and with the permission of the ethical committee of Jorvi hospital. They were subjected to upper gastrointestinal endoscopy with biopsy with a normal forceps including 2–4 specimens from the gastric antrum and 2–4 specimens from the gastric corpus. The Helicobacter‐infection was confirmed histologically by Giemsa staining of the 3–5‐µm sections after formalin fixation and paraffin embedding of the biopsy specimens, as well as by rapid urease test from mucosal specimen (Diabact UBT^R, Orexo, Uppsala, Sweden). The morphologic classification was performed according to Sydney Classification 2,3. All patients underwent an eradication therapy by triple‐therapy with two antibiotics (metronidazole‐clarithromycin) and with a proton pump inhibitor (omeprazole) for one week. Four weeks after termination of the eradication therapy the patients were tested both with ¹³C‐urea breath test and with stool Helicobacter antigen tests. No antibiotics, proton pump inhibitors, H2‐receptor antagonists nor compounds containing bismuth were allowed within 2 weeks before the test.

Gastric mucosal morphology

Nine patients (18%) had mild atrophic gastritis in the gastric antrum. Two patients (4%) had mild, one 2% had moderate and one 2% severe mucosal atrophy in the gastric fundus (scoring: mild, moderate or severe) 3.

Eight patients (16%) had mild intestinal metaplasia in the mucosa of the gastric antrum, one patient (2%) in the fundic mucosa (scoring: mild, moderate, severe) 3.

Forty patients 80% had mild, six patients 12% had moderate and four patients 8% had severe gastritis in the antrum mucosa (scoring: mild, moderate or severe) 3. Correspondingly, 48 patients 96% had mild and 2 patients 4% had severe gastritis in the fundic mucosa.

Testing of Helicobacter pylori infection

The assessment of the effect of eradication therapy was performed by ¹³C‐urea breath test (Pt‐Hepy‐R^R, Orion, Finland), and with monoclonal (Amplified IDEIA HpStA^R, Dako, Denmark) and polyclonal (Premier Platinum HpSA^R, Meridien Bioscience Inc., US) antibody tests for faecal Helicobacter pylori antigen demonstration.

Statistics

In the statistical analysis the assessment of the correlation between mucosal morphology and breath test as well as stool tests, Spearman rank correlation and Pearson's correlation tests were used. A P‐value less than 0.05 was considered significant.

Results

¹³C‐urea breath test and faecal Helicobacter pylori antigen tests

In 16 cases with a positive ¹³C‐urea breath test (Pt‐Hepy‐R ⩾1) the monoclonal test was positive in 15 cases and the polyclonal faecal test in 14 cases. The sensitivity of monoclonal test was 94% and of polyclonal test 88%. In 34 cases with a negative ¹³C‐urea breath test (Pt‐Hepy‐R <1), the monoclonal and the polyclonal faecal tests were negative in 34 and 33 cases, respectively. Thus, the specificity of the monoclonal test was 100% and of the polyclonal test it was 97% (Tables I and II).

Table I. Comparison between ¹³C‐urea breath test and polyclonal faecal Helicobacter pylori antigen test in the detection of Helicobacter colonization.

	Polyclonal faecal test negative	Polyclonal faecal test positive	Total
^13C‐urea breath test negative	33	1	34	Specificity 0.971
^13C‐urea breath test positive	2	14	16	Sensitivity 0.875
Total	35	15	50

Table II. Comparison between ¹³C‐urea breath test and monoclonal faecal Helicobacter pylori antigen test in the detection of Helicobacter colonization.

	Monoclonal faecal test negative	Monoclonal faecal test positive	Total
^13C‐urea breath test negative	34	0	34	Specificity 1.00
^13C‐urea breath test positive	1	15	16	Sensitivity 0.937
Total	35	15	50

The correlation between breath test and the stool test was significant. The Pearson r value for polyclonal test was 0.919 and for monoclonal test 0.794. The Spearman r values were 0.647 and 0.481, respectively (P<0.05 for both) (n = 50).

Gastric mucosal histology and non‐invasive tests

The ¹³C‐urea breath test results as well as the mono‐ and polyclonal stool test results were completely parallel in patients with various forms of mucosal atrophy or intestinal metaplasia in the gastric antrum or fundus. The severity of acute mucosal inflammation did not correlate with the numeric value of ¹³C‐urea breath test and with the values of mono‐ and polyclonal stool tests.

Discussion

The present study showed that the sensitivity and the specificity of faecal Helicobacter pylori antigen testing are comparable to ¹³C‐urea breath test and the correlation between breath test and monoclonal or polyclonal stool test is excellent. Therefore the stool tests can be introduced into clinical practice for the control after Helicobacter pylori eradication therapy. In younger age groups they might even be used for the primary diagnosis of the Helicobacter pylori infection, when applying the test‐and‐treat model of therapy on patients suffering from upper gastrointestinal complaints.

An interesting finding of the present study was the independence of the result of breath test and of stool tests on the presence of mucosal atrophy in the gastric antrum or fundus. Factors that influence the Helicobacter‐colonization are the strain of the bacteria 4 and its ability to produce interleukin‐10 5. On the other hand, the assessment of mucosal atrophy is relatively subjective, and significant inter‐observer variation exists 6. Nevertheless, the present study shows that in case of mild and moderate forms of atrophy, the non‐invasive tests are unaffected, and therefore in practical clinical use of the faecal Helicobacter‐testing the consideration of mucosal atrophy can be ignored, which supports the previous literature 7. Yet, a severe atrophy within the fundic mucosa will presumably be reflected in the ¹³C‐urea test and results in non‐parallel test results with the present non‐invasive stool tests.

There is no previous literature about the correlation between mucosal atrophy in the gastric antrum and stool test. The present study design was not optimal in this regard either. An optimal study design would require morphometric assessment of mucosal atrophy, scored density of Helicobacter pylori colonization and the numeric stool test results.

The eradication therapy is presently performed with two antibiotics combined with a proton pump inhibitor like in the present study. This results in eradication of Helicobacter pylori infection in nearly 90% of cases 8. Even though opinions about the need for the control of the effect of eradication therapy with triple‐therapy diverge, good clinical practice regarding the antibiotic therapy is to control its effect. A successful eradication rate of 68%, like in the present study, demonstrates that the incidence of resistant strains of Helicobacter pylori is increasing, which also indicates control of the effect of eradication therapy. Usually this is performed with ¹³C‐urea breath test or by stool antigen testing with immunoassay. More invasive diagnostic testing with endoscopy and biopsy is required only in case of frank gastric ulceration to exclude possible neoplastic background of the lesion. Presently, the stool antigen testing allows an easy and simple office‐based testing of Helicobacter pylori infection 1. The sensitivity and specificity of stool tests range between 63% and 100% 9–14.

Since 2002 several stool tests, such as culture, polymerase chain reaction (PCR) and immunoassay, have been developed. The culture was primarily used in Gambia for Helicobacter‐infected children 15. Subsequently, various study groups have shown its limitations and it is not recommended for clinical use for Helicobacter pylori infection any more. PCR has been used to detect Helicobacter pylori in various clinical specimens such as gastric biopsies, gastric juice and saliva 16 and in human faeces 17,18. The limitations of the PCR test are false‐negative results due to the presence of PCR inhibitors or to genetic variability, and the false‐positive results due to contamination and non‐specific amplification of human genomic DNA. In addition, the coccoid forms of Helicobacter pylori create a special problem as well as the defective standardization. Therefore, the recent immunoassay stool tests, which were the subject of the present study, are significant improvements in the non‐invasive detection allowing a direct office‐based measurement of Helicobacter pylori.

The common occurrence of active Helicobacter pylori infection and of suspected symptomatic cases indicates the development of less invasive and more practical diagnostic tests than presently available, particularly for younger patient groups. The control of the result of eradication therapy can be performed without control endoscopy if the primary investigation has not revealed any frank mucosal ulcerations. In patients with recurring upper gastrointestinal symptoms and with a suspected relapse, the availability of a non‐invasive test for Helicobacter pylori infection is worthwhile. In addition, due to increasing demands for diagnostic endoscopy, the work‐load ought to be alleviated by developing less invasive diagnostic methods. In this regard the demonstration of stool antigen would be ideal, because it can also be performed office‐based like the ¹³C‐urea breath test, but is more cost‐effective than the other non‐invasive tests 19, when performed 4 weeks after the eradication therapy like in the present series. The sensitivity is known to increase subsequently, therefore there are reports suggesting its performance first until 6–8 weeks after termination of eradication 20–22. Due to its reliability, practicability and cost‐effectiveness, Helicobacter pylori stool testing is presently approved as the primary post‐treatment test after eradication by the American Food and Drug Administration and by the European authorities 8. The test result is sensitive to preceding proton pump inhibitor therapy, but not to H2‐receptor antagonist therapy 23. The unpleasantness of the collection of stool samples may influence the compliance of the test as shown in colorectal cancer screening 24.

In conclusion, the present Helicobacter pylori stool antigen tests either with monoclonal or polyclonal antibodies are equal alternatives to ¹³C‐urea breath test in the detection of Helicobacter pylori infection after eradication therapy regarding their specificity. Regarding the sensitivity, the monoclonal test is slightly better than the polyclonal test, and both are equally insensitive to the mucosal atrophy of gastric antrum.

Acknowledgements

The technical help of Sari Karesvuori, RN, in the practical performance of the study is gratefully acknowledged. This study has been supported by the Clinical Research Institute of the Helsinki University Central Hospital Helsinki, and Astra Co., Espoo, Finland.