Evaluation of a new rapid immunoassay for the detection of Helicobacter pylori in faeces: a prospective pilot study

Transcription

1 Aliment Pharmacol Ther 2005; 21: doi: /j x Evaluation of a new rapid immunoassay for the detection of Helicobacter pylori in faeces: a prospective pilot study L. TREVISANI*, S. SARTORI*, M. R. ROSSI, M.RUINAà, V.MATARESE*,S.GULLINI*& V. ABBASCIANO* *Digestive Endoscopy Centre, Medical Department, S. Anna Hospital, Ferrara; Chemical Chemistry Department, S. Anna Hospital, Ferrara; àpostgraduate School of Gastroenterology, University of Ferrara, Ferrara, Italy Accepted for publication 24 November 2004 SUMMARY Background: Detection of Helicobacter pylori antigen in faeces is a valid method to diagnose H. pylori infection. Presently available stool tests are performed in the laboratory, and diagnostic report is delayed. Aim: To evaluate a new rapid stool test in a pretreatment setting and to compare it with a validated laboratory stool test. Methods: A total of 105 patients underwent gastroscopy with brush cytology, and biopsies for histology and rapid urease test, to assess H. pylori presence. Helicobacter pylori-status was considered positive if at least two tests were positive; negative if all tests were negative; indeterminate if one test was positive and two negative. Stool specimens were tested using either a rapid immunoassay kit (ImmunoCard STAT) or a laboratory enzyme immunoassay kit (Hp StAR). Results: Sixty patients were infected with H. pylori, 44 non-infected, one indeterminate. The sensitivity and specificity of ImmunoCard STAT were 85 and 93%; those of Hp StAR were 88 and 100% (not significant). Conclusions: ImmunoCard STAT seems a reliable method for detecting H. pylori in untreated patients. It could replace laboratory stool tests, as it is easy and can be performed quickly. These characteristics might be a breakthrough for diagnosing H. pylori in the doctor s office. INTRODUCTION A large number of methods have been used to diagnose Helicobacter pylori infection. At present, there is an increasing interest in non-invasive tests, as they are not influenced by sampling error and can profitably replace endoscopy in predicting the diagnosis and determining the management of some types of patients. 1, 2 Detection of H. pylori in faeces by using a non-invasive method is a very interesting diagnostic tool, and several stool antigen tests have been put on the market. A lot of Correspondence to: Dr L. Trevisani, Centro di Endoscopia Digestiva, Azienda Ospedaliera-Universitaria Arcispedale S. Anna, C.so Giovecca 203, Ferrara, Italy. tvl@unife.it reports documented their high diagnostic accuracy, and these tests are considered reliable methods either in pretreatment or in post-treatment H. pylori eradication groups. 3 5 Until now, all available tests are based on an enzyme immunoassay carried out in a laboratory, and such a procedure delays the diagnostic report. Recently, a rapid stool test not requiring laboratory assay has been put on the market, but at present the data about its clinical usefulness are insufficient, 6 8 as it is quite new. The first aim of this prospective pilot study was to evaluate this new test as a predictor of H. pylori status in the pre-treatment setting. The secondary aim was to compare the diagnostic accuracy of this test with a wellknown and validated H. pylori stool test requiring laboratory assay. Ó 2005 Blackwell Publishing Ltd 485

2 486 L. TREVISANI et al. MATERIALS AND METHODS Patients The study involved 105 patients (55 males and 50 females; mean age 57.7 years; range: 23 87), referred to our digestive endoscopy centre for oesophagogastroduodenoscopy (EGD). They had never been treated for H. pylori infection, and underwent diagnostic EGD because of upper gastrointestinal symptoms. Exclusion criteria were the following: age under 18 years; antibiotic or bismuth salts, or proton-pump inhibitors therapy in the last 2 months; chronic use of corticosteroids or non-steroidal anti-inflammatory drugs; prior gastric surgery; bleeding peptic ulcer; severe concomitant diseases; pregnancy or lactation. Informed consent was obtained from each patient, and the study was approved by our local Ethical Committee. Methods During EGD, a brushing sample was obtained by repeatedly brushing antral mucosa with a sterile disposable cytology brush (US Endoscopy Group, Mentor, OH, USA). The brush (Br) was smeared on a slide, air-dried and stained by May Grunwald-Giemsa, as previously reported. 9 Moreover, three biopsies were taken from the antrum, and three from the corpus. Two biopsies (one from the antrum, and one from the corpus) were used for rapid urease test (RUT) (Cp-test, Yamanouchi Pharma, Milan, Italy). The other four biopsies were used for histological examination (H) (haematoxylin eosin and Giemsa stains). RUTs were monitored for colour change up to 6 h at room temperature, after the addition of the gastric samples. The test was considered positive if the colour changed from yellow to red. Cytological and histological assessment of H. pylori presence was carried out by two different doctors, who were blinded each other and unaware of the result of RUT. Helicobacter pylori organisms are curved, spiral or S-shaped, and become intensely violet blue with the staining method used. No single test was used as gold standard to assess H. pyloristatus, which was considered positive if at least two of the three tests were positive; negative if all three tests were negative; and indeterminate if one test was positive and two were negative. All patients were required to deliver a stool specimen the day after EGD. A portion of each sample was tested by using a new rapid immunochromatographic assay commercial kit (ImmunoCard STAT HpSA, Meridian Diagnostic Inc., Cincinnati, OH, USA) (ImmunoCard STAT) for the detection of H. pylori antigens in stool specimens. The remaining portion of each specimen was frozen and stored, and successively tested by using an enzyme immunoassay commercial kit (Amplified IDEA Hp StAR, Dako, Glostrup, DK) (Hp StAR) widely validated in literature, with a sensibility and specificity of about 96% and 97%, respectively, in the pretreatment setting. 10 Only specimens from patients with H. pylori-status assessed as positive or negative were tested. Patients with indeterminate H. pylori-status were excluded from the assay, as diagnostic accuracy of the new test under investigation could be biased in presence of uncertain H. pylori-status. The doctor testing the stool specimens was unaware of the results of RUT, H and Br. The ImmunoCard STAT is an in vitro qualitative procedure based on a lateral flow chromatography technique that detects bacterial antigens using a monoclonal anti-h. pylori antibody. Four drops of diluted stool sample are dispensed to the sample port of the test cassette and the result can be read exactly 5 min later. If only one blue coloured band (control line) appears in the reading window, then H. pylori antigens are absent or below the level of detection (negative result). A positive result is defined if a pink-red band (test line) also appears across the reading window of the device. If either the blue or pink-red band is absent, the result is regarded as invalid. All tests were performed by a single unique observer (L.T.), who evaluated the density of the bands in the immunochromatographic test assay and classified it as positive even if the intensity of the pink-red band (test line) was very weak. The Hp StAR test is an in vitro qualitative procedure for the detection of H. pylori antigens in human stool samples. It is a sandwich-type enzyme immunoassay using immunoassay amplification technology. The wells of the microplate are coated with monoclonal antibodies specific for H. pylori antigens. Supernatant of a faecal suspension as well as horseradish peroxidase (HRP)- labelled monoclonal antibodies (antibody conjugate) are added to the wells in one step. During incubation, H. pylori antigens present in a sample bind to the antibodies on the microplate and to the HRP-labelled antibodies thus forming a sandwich complex. The wells are washed, and a colourless single-component enzyme substrate (tetramethylbenzidine, TMB) is added. Bound HRP oxidizes TMB to a blue-coloured product.

3 ANEWRAPIDH. PYLORI STOOL TEST 487 By adding the stop solution the colour changes to yellow. Assessment of the results was carried out by spectrophotometric determination, reading absorbance at 450 nm within 15 min of adding stop solution. According to the manufacturer s guidelines, an optical density (OD 450 nm ) was defined as a positive and an OD 450 nm < as a negative test result. Estimation of sample size Sample size calculation was performed to obtain a detectable threshold between 10 and 15%, resulting therefore in a sample of about 100 cases. Type I and type II errors were settled at 0.05 and 0.2, respectively. Statistical analysis Sensitivity, specificity, positive and negative predictive values, diagnostic accuracy, and their 95% confidence intervals (CI) of the stool tests were calculated against the H. pylori-status defined as gold standard. The performances of the two stool tests were compared by using Fisher s exact test. P < 0.05 was considered statistically significant. RESULTS Twenty-eight of 105 patients enrolled (26.7%) had active duodenal or gastric ulcer, 26 (24.7%) severe erosive gastritis or duodenitis and 51 (48.6%) functional dyspepsia, mild gastritis or mild duodenitis. Sixty of them (57.14%) were H. pylori-infected, 44 (41.9%) were non-infected. One patient (0.96%) resulted indeterminate, and was excluded from H. pylori stool assays. The correspondence between the results of the ImmunoCard STAT, or Hp StAR, and those of the different biopsy-based methods (H, RUT, Br) used to assess the H. pylori-status is reported in Table 1. In the 104 patients with H. pylori-status positive or negative, ImmunoCard STAT was positive in 54 cases (three false-positives), and negative in 50 (nine falsenegatives), with a sensitivity of 85% (95% CI: 73 93) and specificity of 93.2% (95% CI: 81 99). Hp StAR was positive in 53 cases (all true positives), and negative in 51 (seven false-negatives), with a sensitivity of 88.3% (95% CI: 77 95), and specificity of 100% (95% CI: ). Sensitivity, specificity, predictive values for a positive and negative result and diagnostic accuracy of the two tests are reported in Table 2. No significant Table 1. Correspondence between the results of the ImmunoCard STAT, or Hp StAR, and those of the different biopsy-based methods difference in the performances of the two stool tests was observed, and the concordance between the tests was 95.2% (99 of 104 cases). In no case the two tests had to be repeated owing to equivocal results, although in 15 of 54 cases with positive ImmunoCard STAT the intensity of the pink band was very weak. DISCUSSION ImmunoCard STAT Positive Negative Hp StAR Positive H+ RUT+ Br H+ RUT+ Br) H+ RUT) Br H) RUT+ Br H) RUT) Br) Negative H, histology; RUT, rapid urease test; Br, brush; +, positive; ), negative. The use of non-invasive tests is becoming more and more important, as the diagnosis of H. pylori infections is no longer strictly the domain of the gastroenterologist, but should start at the general practitioner level. 2 Even if serology and 13 C-urea breath test (UBT) are still the non-invasive tests most widely used in clinical practice, the interest in stool tests is progressively increasing. The first commercialized stool test used polyclonal antibodies, and it was put on the market in 1997, supported by just one report presented at the 97th Meeting of the American Society for Microbiology. 11 But then, many reports showing its reliability were published, 10 and it was approved by the FDA for the detection of H. pylori before and after therapy. 12 Subsequently, several others stool tests have been put on the market. The most recent ones are based on monoclonal antibodies, and have shown high performances in both pre- and post-treatment settings, as well as in children use. 5, Until now, the available stool tests have been carried out in laboratory, and the diagnostic report is delayed. Recently, a new rapid stool test has been put on the market (ImmunoCard STAT), and three pilot studies investigating its efficacy reported a sensitivity of % and specificity of %. 6 8 In our study, ImmunoCard STAT correctly

4 488 L. TREVISANI et al. Table 2. Performances of ImmunoCard STAT and Hp StAR tests Sensitivity, % Specificity, % PPV, % NPV, % Diagnostic accuracy, % ImmunoCard STAT 85.0 (73 93) 93.2 (81 99) 94.4 (85 99) 82.0 (69 91) 88.5 (81 94) Hp StAR 88.3 (77 95) 100 (92 100) 100 (93 100) 86.3 (74 94) 93.3 (87 97) CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value. classified 92 of 104 patients, with a global accuracy of 88.5%, confirming the good results previously reported and suggesting that it could be profitably employed in the primary diagnosis of H. pylori infection. Furthermore, we compared this new test with another monoclonal enzyme immunoassay for H. pylori detection in stools (Hp StAR), which has to be performed in laboratory. This test, although rather new, has widely been validated either in adult patients before and after eradication therapy, or in children. 5, In our series, the concordance between the two stool tests was 95.2%, and no significant difference was observed, although Hp StAR seemed slightly more effective than ImmunoCard STAT. Our sample size was fairly small and able to identify as statistically significant only differences of 10 15% in the performances of the two tests, with an a-error of 5% and b-error of 20%. The sample size was planned in such a way, as just a remarkable difference in the performances of the two tests could justify the use of a test more time-consuming and requiring a laboratory assay, rather than a near-patient test to be easily and quickly used in the doctor s daily practice. As recommended by the Maastricht Consensus Report, a test and treat approach should be offered to adult young dyspeptic patients without alarm symptoms. 4 To this aim, the availability of non-invasive, reliable and low-cost tests is quite important. The price comparison between stool tests and UBT depends on the individual country. However, everywhere UBT is considerably more expensive than a stool test ( US$ vs US$ in the USA), 8, 12 which is consequently considered a highly cost-effective method for the diagnosis of H. pylori infection. 10 The cost of a traditional stool test includes the time-consumed by the technician for sample processing, i.e. about 3 h for 46 assays. 16 The availability of a rapid test allows to evade this item, and the cost of the test is just represented by the cost of the kit. At present, in Italy, the cost of the ImmunoCard STAT kit (20 assays) is 520. Consequently, the cost for each assay is 26. In conclusion, this novel rapid immunochromatographic H. pylori stool antigen test seems to have good diagnostic accuracy in the pre-treatment setting, and could represent a valid alternative to the traditional stool tests that have to be performed in laboratory. It is cheap, easy and fast to be performed, and these characteristics might be a breakthrough for diagnosing H. pylori infection in the doctor s office. Further studies should be planned to confirm our results on larger series, and to investigate the reliability of the test in the children and in the posttreatment setting. REFERENCES 1 American Gastroenterological Association Medical Position Statement. Evaluation of dyspepsia. Gastroenterology 1998; 114: European Helicobacter pylori Study Group. Current European concepts in the management of Helicobacter pylori infection. The Maastricht Consensus Report. Gut 1997; 41: Lehmann FS, Beglinger C. Current role of Helicobacter pylori stool tests. Digestion 2003; 68: Malfertheiner P, Megraud F, O Morain C, et al. Current concepts in the management of Helicobacter pylori infection. The Maastricht Consensus Report. Aliment Pharmacol Ther 2002; 16: Weingart V, Russmann H, Koletzko S, et al. Sensitivity of a novel stool antigen test for detection of Helicobacter pylori in adult outpatients before and after eradication therapy. J Clin Microbiol 2004; 42: Gatta L, Perna F, Ricci C, et al. A rapid immunochromatographic assay for Helicobacter pylori in stool before and after treatment. Aliment Pharmacol Ther 2004; 20: Li YH, Guo H, Zhang PB, et al. Clinical value of Helicobacter pylori stool antigen test, ImmunoCard STAT HpSA, for detecting H. pylori infection. World J Gastroenterol 2004; 10: Wu IC, Ke HL, Lo YC, et al. Evaluation of a newly developed office-based stool test for detecting Helicobacter pylori: an extensive pilot study. Hepatogastroenterology 2003; 50:

5 ANEWRAPIDH. PYLORI STOOL TEST Caselli M, Trevisani L, Pazzi P, et al. Suggestions for the rapid diagnosis of Campylobacter pylori infection in endoscopic settings. Endoscopy 1989; 21: Gisbert JP, Pajares JM. Stool antigen test for the diagnosis of Helicobacter pylori infection: a systematic review. Helicobacter 2004; 9: Kozak K, Larka C, Nickol A, et al. Detection of Helicobacter pylori Antigen in Stool Specimens using a Novel Enzyme Immunoassay. American Society for Microbiology Miami Beach, USA, 4 8 May, 1997: C Vakil N, Affi A, Robinson J, et al. Prospective blinded trial of a fecal antigen test for the detection of Helicobacter pylori infection. Am J Gastroenterol 2000; 95: Agha-Amiri K, Peitz U, Mainz D, et al. A novel immunoassay based on monoclonal antibodies for the detection of Helicobacter pylori antigens in human stool. Z Gastroenterol 2001; 39: 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: Makristathis A, Barousch W, Pasching E, et al. Two enzyme immunoassay and PCR for detection of Helicobacter pylori in stool specimens from pediatric patients before and after eradication therapy. J Clin Microbiol 2000; 38: Trevisani L, Sartori S, Ruina M, et al. Helicobacter pylori stool antigen test. Clinical evaluation and cost analysis of a new enzyme immunoassay. Dig Dis Sci 1999; 44: