Archives of Medical Research 37 (2006) 123–128

ORIGINAL ARTICLE

Diagnostic Utility of Invasive Tests and Serology for the Diagnosis ofHelicobacter pylori Infection in Different Clinical Presentations

Jaime Raul Zuniga-Noriega,a Francisco Javier Bosques-Padilla,a Guillermo Ignacio Perez-Perez,b

Rolando Tijerina-Menchaca,c Juan Pablo Flores-Gutierrez,d Hector Jesus Maldonado Garza,a

and Elvira Garza-Gonzalezc

aServicio de Gastroenterologıa del Hospital Universitario Dr. Jose Eleuterio Gonzalez, Monterrey, Nuevo Leon, MexicobDepartments of Medicine and Microbiology, New York University School of Medicine, New York, NY

cDepartamento de Microbiologıa de la Facultad de Medicina, Universidad Autonoma de Nuevo Leon, Monterrey, Nuevo Leon, MexicodLaboratorio de Anatomıa Patologica, Hospital Universitario Dr. Jose Eleuterio Gonzalez, Monterrey, Nuevo Leon, Mexico

Received for publication August 18, 2004; accepted April 27, 2005 (ARCMED-D-04-00060).

Background. Invasive and noninvasive tests are used for the diagnosis of Helicobacterpylori infection. The aim of this study was to determine the diagnostic utility of rapidurease test (RUT), culture, histology and serology for the diagnosis of H. pylori inpatients with different clinical presentations.

Methods. We studied 527 consecutive patients (mean age, 52.5 years; F:M, 1.3; agerange 15–89 years) enrolled at the Hospital Universitario, Universidad Autonoma deNuevo Leon. Patients had gastric cancer (GC, 9.1%), non-ulcer dyspepsia (NUD, 81.4%),or peptic ulcer disease (PUD, 9.1%). The infection by H. pylori was determined byhistology, rapid urease test, culture, and serology. Patients were determined as infectedwith H. pylori if at least a) two invasive tests were positive and b) two tests were positive(invasive or non-invasive). Diagnostic utility was calculated for each assay.

Results. Prevalence of infection in the whole studied population was 50.9%. In NUDpatients the prevalence was 51.3%, in PUD patients 58.3%, and in GC patients 39.6%.When we used the first diagnostic criteria, for the whole studied population, the RUT wasthe most reliable test, followed by the culture. Histology had the best sensitivity for thewhole studied population and NUD patients and RUT had the best sensitivity value forthe GC patients. In the whole studied population, NUD and GC patients, RUT and culturehad the best specificity, accuracy and PPV. For PUD patients, serology had the bestperformance. When we used the second diagnostic criteria, histology and serology hada better performance compared with the results obtained with the first diagnostic criteria.

Conclusions. Diagnostic utility of the tests varies according to the clinical presentations,which should be considered in the selection of the diagnostic test for the detection ofH. pylori. � 2006 IMSS. Published by Elsevier Inc.

Key Words: Helicobacter, Serology, Diagnosis, Diagnostic, RUT, Culture.

Introduction

Infection by Helicobacter pylori is associated with thedevelopment of superficial chronic gastritis, gastric or

Address reprint requests to: Elvira Garza-Gonzalez, Departamento de

Microbiologıa de la Facultad de Medicina, Universidad Autonoma de

Nuevo Leon, Av. Madero y Dr. Aguirre s/n, Colonia Mitras Centro, 64460,

Monterrey, Nuevo Leon, Mexico; E-mail: elvira_garza_gzz@yahoo.com

0188-4409/06 $–see front matter. Copyright � 2006 IMSS. Published by Elsdoi: 10.1016/j.arcmed.2005.04.020

duodenal ulcers, and distal gastric cancer (GC). Accordingto the evidence, H. pylori has been recognized by the WorldHealth Organization (WHO) as a class I gastric carcinogen(1–3).

Several diagnostic methods for detecting H. pyloriare available. These tests can be divided into invasivemethods [rapid urease test (RUT), histology, culture andpolymerase chain reaction (PCR)] that require endoscopy;and noninvasive tests, mainly urea breath test (UBT),

evier Inc.

124 Zuniga-Noriega et al./Archives of Medical Research 37 (2006) 123–128

antigen detection in stool and serologic detection ofantibodies (4–7).

The selection of the diagnostic test should considerseveral aspects such as sensitivity, specificity of the diag-nostic test and the clinical situation, their present complaints,and past medical history. Also, economic evaluations thatassess costs should be done (4–7).

Different diagnostic utilities have been reported for thesemethods and the variations have been attributed mainly tocharacteristics of the colonization of the bacteria such asthe patchy distribution of the microorganism in the mucosalayer of the stomach, the viability of the bacteria, the pre-sence of blood and the histologic characteristics of stomachepithelia as have been described that the pre-malignant andmalignant conditions are not suitable for the growth ofH. pylori (1,3,8). These characteristics are different amongthe different clinical presentations H. pylori-related (3,5,6–10), and for this reason it is necessary to evaluate thediagnostic utility of each test according to clinicalpresentations. The aim of this study was to determine thediagnostic utility of RUT, culture, histology and serology orthe diagnosis of H. pylori in several clinical presentations.

Materials and Methods

Study Population

We studied 527 consecutive patients (mean age, 52.5 years;F:M, 1.3; age range, 15–89 years, median, 54 years) whohad GC (n5 48, mean age, 58.7 years; F:M, 0.7; age range,22–84 years; median, 60 years), non-ulcer dyspepsia (NUD)(n5 431, mean age, 50.2 years, F:M, 1.8; age range, 15–92years; median, 50 years; and peptic ulcer disease (PUD)(n 5 48, mean age, 60.5 years; F:M, 0.4, age range, 17–91years; median, 60 years).

We excluded patients who had received antibiotics orproton pump inhibitors during the 4 weeks prior toendoscopy. Patients were enrolled at the Hospital Uni-versitario Dr. Jose Eleuterio Gonzalez, Universidad Auto-noma de Nuevo Leon. All invited patients accepted toparticipate and signed an informed consent approved by thelocal ethics committee.

Histopathologic Examinations

During the endoscopic procedure, eight biopsy specimenswere obtained for detection of the bacteria and histologicalevaluation, two from the lower curvature, two from thegreater curvature, two from the incisura angularis, and twofrom the pre-pyloric region. All sample biopsies were fixedin 10% formalin, paraffin embedded and multiple 4-mm-thick histological sections were stained with hematoxylin-eosin for histopathologic evaluation by a specializedpathologist unaware of the other results.

Culture

Four biopsy specimens (two from antrum and two fromcorpus) were placed directly in Stuart’s transport media andcultured according to standard procedures (9).

Rapid Urease Test (RUT)

RUT was performed by a non-commercial validated test(10). One single antrum biopsy specimen was placed ina test vial, incubated at 37�C, and read after 24 h. A colorchange from orange to pink indicated a positive result.

Serology

Enzyme-linked immunosorbent assay (ELISA) was used tostudy the presence of IgG antibodies to whole cell (WC) inpatient’s serum samples as previously described (11).

Clinical Classification

The presence of GC or PUD was defined according tohistological and endoscopic findings. The presence of NUDwas defined when no organic disease was found and thepatient presented upper abdominal or epigastric pain,symptoms related to meals, and other symptoms such asheartburn, nausea, or vomiting (12).

Diagnosis of Infection by H. pylori

In the whole studied population and in each clinicalpresentation we used the next criteria: true positive (TP)were patients with at least two invasive tests positive, truenegative (TN) were patients with all invasive diagnostictests negative, false positive (FP) were patients with theanalyzed test positive and all invasive tests negative and thefalse negative (FN) were patients with the tests analyzednegative and at least two invasive tests positive.

In addition, we used a second criteria to define thepresence of H. pylori infection: true positive (TP) werepatients with at least two positive tests whether they wereinvasive or non-invasive, true negative (TN) were patientswith all the diagnostic tests negative, false positive (FP)were patients with the analyzed diagnostic test positive butall the other tests negative, and false negative (FN) werepatients with the test under the analysis negative but at leasttwo other diagnostic tests were positive.

Statistics

With the TP, TN, FP, and FN values defined above, wedetermined the sensitivity and specificity for each di-agnostic test in each clinical condition. Also, sensitivity andspecificity were combined into a single parameter, thelikelihood ratio (LR): the odds (likelihood) of beinginfected if the test result was positive (LR1) and the odds

125Diagnosis of H. pylori Infection in Different Clinical Presentations

of being infected if the test result is negative (LR2).Additionally, the probability that a patient was infected givena positive result (positive predictive value, PPV), theprobability that a patient was not infected given a negativeresult (negative predictive value, NPV), and the accuracywere calculated as well as the corresponding 95% confidenceintervals for all tests.

Results

Prevalence of Infection by H. pylori in the Entire StudiedPopulation and in the Different Clinical Presentations

The prevalence of infection by H. pylori was defined by thepositive result of at least two invasive tests positive andaccording to these results, the prevalence of infection in thewhole studied population was 50.9%. In NUD patients theprevalence was 51.3%, in PUD patients 58.3%, and in GCpatients 39.6%.

Results of the Diagnostic Tests in Each ClinicalPresentation and in the Entire Studied Population

Fourteen different combinations of results of the diagnostictests were obtained (Table 1). Among the NUD and PUDpatients, the majority were positive by all four diagnostictests. Among the GC patients, the most frequent combina-tion was histology and serology positive and both RUT andculture negative. With these results, the values of TP, TN,FP and FN were obtained for each diagnostic test in eachclinical condition and in the whole studied population.

Assessment of the Sensitivity, Specificity, Accuracy, PPV,and NPV in the Entire Studied Population

According to the results obtained for each test, sensitivity,specificity, accuracy, PPV, and NPV were assessed for each

diagnostic test for the whole studied population (Figure 1).According to these results, RUT is the most reliable test.The histology showed the best sensitivity value comparableto serology and RUT and culture had the best specificity,accuracy, and PPV.

Assessment of the Sensitivity, Specificity, Accuracy, PPV,and NPV in the Different Clinical Presentations

Diagnostic utility was calculated for each clinical pre-sentation (Figure 1). For NUD patients, histology had thebest sensitivity and RUT and culture had the bestspecificity, accuracy and PPV (Figure 1). For GC patients,RUT had the best sensitivity, RUT and culture had the bestspecificity, accuracy and PPV. For PUD patients, serologyhad the best performance for sensitivity, specificity,accuracy and PPV.

LR1 and LR2 values were obtained for histology, RUTand serology in each population studied. The valuesobtained for LR1 and LR2 in different clinical conditionswere obtained for histology, RUT and serology (Table 2).The calculation of the likelihood ratio was not possible forculture for mathematical reasons (zero value in thedenominator).

We observed that for all populations studied, RUT hadthe best LR1 and that all three tests had an acceptableLR2 value for the whole studied population and for NUDpatients. For both gastric cancer and PUD patients, theLR2 values were higher than for the whole population andNUD patients.

Diagnostic utility of histology for each clinical pre-sentation test when the criteria of two test positive wasused. We also analyzed the results considering a patient asinfected with H. pylori by the positive result of two tests,independently if they were invasive or not (Figure 2).

Table 1. Results of the diagnostic test in each clinical presentation and in the studied population

Histology RUT Culture Serology NUD n (%) PUD n (%) GC n (%) Total n (%)

1 1 1 1 162 (37.6) 19 (39.6) 10 (20.8) 191 (36.3)

2 2 2 2 60 (13.9) 2 (4.17) 6 (12.5) 68 (12.9)

1 2 2 2 54 (12.5) 5 (10.4) 6 (12.5) 65 (12.3)

1 1 2 2 5 (1.1) 0 (0) 0 (0) 5 (1)

1 1 1 2 2 (0.4) 1 (2.0) 3 (6.2) 6 (1.1)

2 1 2 2 3 (0.7) 0 (0) 0 (0) 3 (0.6)

2 1 1 2 2 (0.4) 0 (0) 0 (0) 2 (0.4)

2 1 1 1 1 (0.2) 1 (2.0) 2 (4.1) 4 (0.8)

2 2 2 1 15 (3.4) 0 (0) 5 (10.4) 20 (3.8)

1 2 2 1 78 (18.1) 12 (25) 11 (22.9) 101 (19.2)

1 1 2 1 37 (8.5) 4 (8.3) 4 (8.3) 45 (8.5)

1 2 1 1 10 (2.3) 3 (6.25) 0 (0) 13 (2.5)

1 2 1 2 2 (0.4) 0 (0) 0 (0) 2 (0.4)

2 1 2 1 0 (0) 1 (2.0) 1 (2.0) 2 (0.4)

431 (100) 48 (100) 48 (100) 527

RUT, rapid urease test; NUD, non-ulcer dyspepsia; PUD, peptic ulcer disease; GC, gastric cancer.

126 Zuniga-Noriega et al./Archives of Medical Research 37 (2006) 123–128

A

0

20

40

60

80

100

Histology

B

0

20

40

60

80

100

C

0

20

40

60

80

100

D

0

20

40

60

80

100

RUT Culture Serology

Histology RUT Culture Serology Histology RUT Culture Serology

Histology RUT Culture Serology

Figure 1. Sensitivity, specificity, accuracy, positive predictive value and negative predictive value with its corresponding 95% confidence intervals for each

diagnostic test in the whole population studied and according to clinical presentations. (A) For the whole population, (B) patients with non-ulcer dyspepsia,

(C) patients with gastric cancer, and (D) patients with peptic ulcer disease.

Sensitivity

Specificity

Accuracy

Positive predictive value

Negative predictive value

Data not shown were not calculated for mathematical reasons.

According to these results, histology and serology showeda better performance, particularly for specificity.

Discussion

H. pylori colonizes the gastric mucosa and induces aninflammatory response that can lead to the formation ofulcers or to the development of preneoplastic conditions,which paradoxically no longer allow the growth of thebacteria (1–4). The transformation of gastric epithelia arisesdifferences in the load, distribution and viability of

microorganisms, and the utility of diagnostic tests used todetect H. pylori are closely related with these differences;therefore, it is expected that the utility of the diverse testsvaries among clinical presentations. In this study weobtained different diagnostic utilities in the entire studiedpopulation (with indication of an upper endoscopy) and indifferent clinical presentations (PUD, NUD, and GC),which suggests that the clinical condition should beconsidered when selecting the test to use.

Accurate detection of the organism is essential forproper patient management and particularly for the

Table 2. Positive likelihood ratio (LR1) and negative likelihood ratios (LR2) for histology, RUT and serology in each population studied

Histology RUT Serology

Entire studied population LR1 (95% CI) 1.5 (1.4–1.6) 17.3 (7.4–40.6) 2.2 (1.8–2.6)

LR2 (95% CI) 0.09 (0.04–0.19) 0.07 (0.05–0.12) 0.1 (0.06–0.17)

Non-ulcer dyspepsia patients LR1 (95% CI) 1.5 (1.4–1.7) 24.2 (8–73) 2.14 (1.73–2.65)

LR2 (95% CI) 0.05 (0.02–0.15) 0.07 (0.04–0.12) 0.09 (0.05–0.16)

Gastric cancer patients LR1 (95% CI) 1.43 (1.0–2.0) 12 (1.8–78.4) 1.9 (0.94–3.6)

LR2 (95% CI) 0.34 (0.09–1.34) ND 0.29 (0.09–0.93)

Peptic-ulcer disease patients LR1 (95% CI) 1.06 (0.89–1.3) 2.6 (0.52–13) ND

LR2 (95% CI) 0.45 (0.04–4.6) 0.2 (0.05–0.74) ND

RUT, rapid urease test; ND, not determined.

127Diagnosis of H. pylori Infection in Different Clinical Presentations

A

0

20

40

60

80

100

B

0

20

40

60

80

100

C

0

20

40

60

80

100

D

0

20

40

60

80

100

Histology RUT Culture SerologyHistology RUT Culture Serology

Histology RUT Culture Serology Histology RUT Culture Serology

Figure 2. Sensitivity, specificity, accuracy, positive predictive value and negative predictive value with its corresponding 95% confidence intervals for each

diagnostic test in the entire population studied and according to clinical presentations when the diagnosis was performed by the positive result of two tests,

whether they were invasive or not. (A) For the whole population, (B) patients with non-ulcer dyspepsia, (C) patients with gastric cancer, and (D) patients with

peptic ulcer disease.

Sensitivity

Specificity

Accuracy

Positive predictive value

Negative predictive value

Data not shown were not calculated for mathematical reasons.

eradication of the bacteria following treatment. This studyshows the diagnostic utility for invasive and non-invasivetests in the initial diagnosis of infection and does not reflectthe diagnostic utility to detect eradication of the infection.

For initial detection of the infection by H. pylori, it isgenerally considered whether or not the endoscopy isindicated, and this is the key factor to decide to use invasiveor non-invasive tests. When an upper endoscopy should beperformed and invasive tests can be done to detectH. pylori, culture and RUT seem to be good options accord-ing to the specificity, accuracy, and PPV values for theentire studied population, NUD, and GC patients. Betweenthese tests, the culture is not widely used because thebacteria grow very slowly in culture media and is moreexpensive than other options, is technically difficult, andrequires strict conditions (3,5,9). In addition, according tothe results of this study, the culture had the lowestsensitivity for the whole studied population and for eachclinical presentation. This diagnostic assay has several

important and unique advantages such as the highestspecificity, antimicrobial therapy can be planned on thebasis of the organism’s susceptibility, and virulence factorscan be investigated by molecular methods.

A particularly interesting result in this work was thatthere were 130 patients with only positive histology.Although it is possible that those patients represented falsepositive results, this phenomenon could be explained by thedifference in the number of samples obtained for each test.As an example, for CLO test we used only one biopsy, butfor the histological analysis we obtained eight biopsies(four from corpus, two from antrum, and two from incisuraangularis) and the number of biopsies for culture were four(two from antrum and two from corpus). Furthermore, thisexplanation is supported by the improved results observedfor histology when we analyzed the data with the secondcriteria, in which patients were considered as infected whentwo tests were positive, whether they were invasive or not.The results of this study underscore the importance of the

128 Zuniga-Noriega et al./Archives of Medical Research 37 (2006) 123–128

validation of the diagnostic tests not only in particularpopulations and clinical conditions but also considering thetopography of infection.

Endoscopy is expensive, unpleasant for patients, andcarries a small but definite risk for complications. Thus, theuse of non-invasive tests is becoming more important in thediagnosis of H. pylori infection. According to the results ofthis study, serology had a good performance in all clinicalpresentations, including GC patients in whom the diagnosisof infection by H. pylori could have little therapeuticrelevance but significant epidemiological interest.

Besides the diagnostic utility of tests, factors such asaccessibility, process time, cost, and knowledge of theclinical condition must be considered when selecting thediagnostic tests for the detection of H. pylori.

In this work we evaluated methods that requireendoscopy for their process (RUT, culture, histology) anda non-invasive method (serology) in three clinical presenta-tions associated with the infection by H. pylori (NUD, PUD,and GC). It should be noted that several other tests notconsidered in this study are available and widely used for thediagnosis of H. pylori infection, such as the UBT, the stoolantigen test and the PCR (3,5,7,13). Particular interest relieson UBT because this test is currently FDA approved and hasshown a high sensitivity and specificity (94.7 and 95.7%,respectively) (14). This test does not have the sampling errorof biopsy-based tests, and in contrast to serology, falsepositive results are uncommon and false negative tests mayoccur mainly related to some medications.

Sensitivity and specificity are technical parameters ofdiagnostic testing performance and have important impli-cations for screening and clinical practice guidelines;however, they are less relevant in the typical clinicalsetting because the clinician does not know whether or notthe patient has the disease. The PPV, NPV, LR1 and LR2give a better idea of the performance of a specific test.

For the whole population studied, RUT is the mostreliable test, followed by the culture. From a diagnosticpoint of view, it seems that the combination of serology andRUT should be optimal for the diagnosis of H. pyloriinfection; however, the limitation of RUT as an invasive testshould always be considered, as well as the clear

convenience of the histological analysis for the definitionof the clinical condition of the patient.

Unlike the published information that emphasizes theyield of these tests without taking into account the clinicalstatus, we found that the diagnostic utility of each testvaries according to the clinical presentation, which shouldbe taken into account when deciding the test to use to detectthe presence of the bacterium.

References1. Suerbaum S, Michetti P. Helicobacter pylori infection. N Engl J Med

2002;347:1175–1186.

2. NIH consensus statement. Helicobacter pylori in peptic ulcer disease.

JAMA 1994;272:65–69.

3. Dunn BE, Cohen H, Blaser MJ. Helicobacter pylori. Clin Microbiol

Rev 1997;10:720–741.

4. de Boer WA, de Laat L, Mergaud F. Diagnosis of Helicobacter pylori

infection. Curr Opin Gastroenterol 2000;16(Suppl 1):S5–S10.

5. Mitchell H, Megraud F. Epidemiology and diagnosis of Helicobacterpylori infection. Helicobacter 2002;7(Suppl 1):8–16.

6. Hoang TT, Wheeldon TU, Bengtsson C, Phung DC, Sorberg M,

Granstrom M. Enzyme-linked immunosorbent assay for Helicobacterpylori needs adjustment for the population investigated. J Clin

Microbiol 2004;42:627–630.

7. Gatta L, Ricci C, Tampieri A, Vaira D. Non-invasive techniques for the

diagnosis of Helicobacter pylori infection. Clin Microbiol Infect 2003;

9:489–496.

8. Lee JM, Breslin NP, Fallon C, O’Morain CA. Rapid urease tests lack

sensitivity in Helicobacter pylori diagnosis when peptic ulcer disease

presents with bleeding. Am J Gastroenterol 2000;95:1166–1170.

9. Perez-Perez GI. Accurate diagnosis of Helicobacter pylori. Culture,

including transport. Gastroenterol Clin North Am 2000;29:879–884.

10. Flores-Orta D, Bosques-Padilla F, Gomez-Leija G, Frederick F.

Comparative study of rapid urease test (Hazell test) vs CLO-test in the

diagnosis of Helicobacter pylori infection. Gut 1997;41(Suppl 3):

A160.

11. Perez-Perez GI, Dworkin BM, Chodos JE, Blaser MJ. Campylobacterpylori antibodies in humans. Ann Intern Med 1998;109:11–17.

12. Jaakkimainen RL, Boyle E, Tudiver F. Is Helicobacter pylori

associated with non-ulcer dyspepsia and will eradication improve

symptoms? A meta-analysis. BMJ 1999;319:1040–1044.

13. Smith SI, Oyedeji KS, Arigbabu AO, Cantet F, Megraud F, Ojo OO,

Uwaifo AO, Otegbayo JA, Ola SO, Coker AO. Comparison of three

PCR methods for detection of Helicobacter pylori DNA and detection

of cagA gene in gastric biopsy specimens. World J Gastroenterol 2004;

10:1958–1960.

14. Vaira D, Vakil N. Blood, urine, stool, breath, money, and Helicobacter

pylori. Gut 2001;48:287–289.