Among the non invasive tests UBT has excellent sensitivity and specificity. However, UBT is a highly expensive test and requires trained staff which makes it a less acceptable choice in present population. Serology based tests are poor differentiators of current and past infection and often yields misleading results in areas of high prevalence [6]. Thus, pathogen-specific stool antigen tests that detect active infection are a valid choice of non invasive tests to detect H.pylori infection. Both European and Japanese guidelines endorsed the use of SATs using monoclonal antibodies for primary diagnosis as well as for the assessment of eradication therapy [7].
Stool antigen tests are available in two formats: ELISA and Immunochromatography (ICT). Faecal H.pylori antigen tests based on immunochromatographic reactions are more widely used nowadays due to their ease of use and cost benefits in addition to their rapidity. As the accuracy of the test may vary in different populations due to the difference in the antigenicity of H.pylori strains, a local validation of SAT is essential before use as a diagnostic test [8]. There are only few published data from India regarding the validation of these newer stool antigen test kits [9].
The present study conducted at Government Medical College, Thrissur, Kerala, India, was an attempt to validate a locally available rapid SAT for diagnosing H.pylori infection among dyspeptic patients. The H.pylori infection among dyspeptic patients is diagnosed primarily by endoscopic biopsy and histopathological examination in present institution. The availability of locally validated SAT can not only reduce the endoscopy workload but also provide an acceptable non invasive diagnostic method.
Materials and Methods
A prospective diagnostic validation study was carried out among randomly selected 113 patients above 14 years of age who were undergoing upper gastrointestinal endoscopy for evaluation of dyspepsia at the Department of Gastroenterology, Government Medical College, Thrissur, kerala, India. The study was conducted over a period of one year from March 2015 to February 2016 after Institutional Ethics Committee approval (Reg. No 170/2). The sociodemographic and clinical data were collected after a written informed consent. Patients who were treated with antibiotics or proton-pump inhibitors within two weeks prior to endoscopy or with history of gastric surgery were excluded from the study.
During endoscopy three gastric biopsy samples (two from antrum and one from body) were collected from each patient. Sampling from antrum as well as corpus biopsy from greater curvature was done to avoid false negative results due to patchy distribution of H.pylori in the stomach. One antral biopsy sample was used for RUT and other two were used for histopathology examination. The RUT was done using 0.5 mL of RUT broth (Himedia M1828). The test was considered as positive when the colour of the medium changed from yellow to pink on incubation at 37°c for maximum upto 18 hours [10]. Histopathology examination was carried out in the Department of Pathology using Haematoxylin and Eosin stain and Giemsa stain as per standard protocols.
Stool Antigen Test
Stool samples were collected in sterile wide mouthed containers and rapid SAT was done immediately using Epituub® fecal H.pylori antigen test kit, a commercially available kit based on ICT assay according to manufacturer’s (Epitope Diagnostics, Inc. Sandiego, USA) instructions. Stool samples were transferred to the collection tube containing extraction solution to extract H.pylori antigens from faeces. The sampling tube was mixed vigorously to ensure a good liquid suspension and then the sampling tube was positioned upside down in vertical allowing the stool particles to sediment for about one minute. Then the test strip was removed from the sealed foil pouch and screwed in a vertical position into the sampling tube by breaking into the bottom seal of the sampling tube. Thus, the solution was allowed to flow into the bottom space of the test strip. The test result was observed at 10 minutes. If two red/pink coloured bands were visible at the test area and control area within 10 minutes, the test result was considered positive and valid. If test area has no coloured band and the control area displays a red/pink coloured band, the test result was negative. If a coloured band was not formed in the control area regardless there is any band in the test area, the test result was considered invalid.
The gold standard diagnostic criteria for H. spylori infection in the present study was defined as a positive test result for both rapid urease test and histopathology examination. Positive test result for only one of these tests was considered as negative.
Statistical Analysis
Data were analysed by descriptive analysis using Epiinfo software. Quantitative variables were expressed as means±standard deviation while qualitative variables were expressed as percentages. Chi-square test was used to compare categorical data. The test performance was assessed by determining sensitivity, specificity, positive predictive value, negative predictive value and diagnostic accuracy.
Results
The age of patients ranged from 15 to 79 years and mean age±SD was 45.8±14.7 years. Out of the 113 patients with dyspepsia, 54% (61/113) were females and 46% (52/113) were males. Based on the on the gold standard diagnostic criteria, 35 (31%) patients were diagnosed as H.pylori infected and 78 (69%) as uninfected. Histopathology showed H.pylori positivity in 61% of the cases (69/113) where as RUT showed positivity in only 38% (43/113) of the cases. Of 113 patients, 41% (46/113) were rapid SAT positive.
Majority of the patients studied belonged to the age group 50-59 years with maximum H.pylori positivity. H.pylori positivity was also more in females, 23 (37.7%) as compared to males, 12 (23%). There was no statistically significant association between female sex and H.pylori positivity. [Table/Fig-1] depicts the age distribution and [Table/Fig-2] shows the age and sex characteristics of the study population.
Age distribution among the study population.
Age Group | Frequency (Percentage) | H.pylori infected (Percentage) |
---|
15-19 | 3 (2.65%) | 0 |
20-29 | 17 (15.04%) | 5 (29.4%) |
30-39 | 17 (15.04%) | 4 (23.5%) |
40-49 | 24 (21.24%) | 8 (33.3%) |
50-59 | 27 (23.89%) | 10 (37%) |
60-69 | 22 (19.47%) | 7 (31.8%) |
70-79 | 3 (2.65%) | 1 (33.3%) |
Total | 113 (100.00%) | 35 |
Age and sex characteristics of the study population and their association with H.pylori infection.
| | No of patients, n (%) | H.pylori infection n (%) | Chi-square | p-value |
---|
Age | ≤45 years | 55 (48.7) | 15 (27.2) | 0.6864 | 0.4 |
>45 years | 58 (51.3) | 20 (34.4) |
Sex | Males | 52 (46) | 12 (23) | 2.8 | 0.09 |
Females | 61 (54) | 23 (37.7) |
Endoscopic abnormalities were observed in 75 out of 113 patients (66.3%). Gastric ulcers were present in 11 (9.73%) patients where as 9 (8%) patients were suffering from duodenal ulcer. Gastritis was observed in 18 patients. Other endoscopic findings were hiatus hernia, proximal gastropathy, polyp, oesophagitis, telengectasia etc. One patient had carcinoma stomach. [Table/Fig-3] shows the characterisation of H.pylori infected population based on endoscopic diagnosis. One patient was having both hiatus hernia and gastric ulcer. Another patient was having both proximal gastropathy and hiatus hernia. Patients with other findings were not having H.pylori infection.
Characterization of H.pylori infected population based on endoscopic diagnosis.
| Endoscopic finding | Total No n (%) | H.pylori Infection No (%) | X2 | p-value |
---|
0 | Normal endoscopy | 38 (33.63) | 12 (31.57) | 0.0098 | 0.9 |
1 | Gastritis | 18 (16) | 5 (27.7) | 0.1023 | 0.74 |
2 | Hiatus hernia | 13 (11.5) | 4 (30.8) | 0.0003 | 0.986 |
3 | Gastric ulcer | 11 (9.73) | 6 (54.5) | 3.16 | 0.07 |
4 | Proximal gastropathy | 10 (8.8) | 7 (70) | 7.815 | 0.005 |
5 | Duodenal ulcer | 9 (8) | 3 (33.3) | 0.0255 | 0.87 |
Note: One patient was having both hiatus hernia and gastric ulcer. Another patient was having both proximal gastropathy and hiatus hernia
Only proximal gastropathy was statistically associated with H.pylori infection (p-value=0.005). Histopathological examination of gastric biopsies obtained from 113 patients showed chronic gastritis in 88 patients, H.pylori was demonstrated by staining in 68 out of these 88 patients and this association was significant (p<0.001). Based on the diagnostic criteria 34 out of 88 chronic gastritis patients (38.6%) were H.pylori infected (p<0.001) [Table/Fig-4,5] respectively.
Correlation of H.pylori infection and histopathological diagnosis.
Histopathological diagnosis | no | H.pylori infected No (%) | Chi-square | p-value |
---|
Chronic gastritis | 88 | 34 (38.6) | 10.92 | 0.001 |
Polyp | 2 | 0 | | |
Carcinoma | 1 | 1 | | |
Erosive gastritis | 3 | 0 | | |
No pathology | 19 | 0 | | |
Correlation of H.pylori staining and histopathological diagnosis.
Histopathological diagnosis | No | H.pylori positivity by staining | Chi-square | p-value |
---|
Chronic gastritis | 88 | 68 | 43.96 | 0.001 |
Others | 6 | 1 | | |
No pathology | 19 | 0 | | |
Sensitivity, specificity, predictive values and accuracy of the stool antigen test kit were calculated in relation to the diagnostic criteria. The rapid stool antigen test detected H.pylori antigen in 31 of the 35 H.pylori-infected patients (sensitivity 88.5%; 95% Confidence Interval (CI): 85.4-91.6%), and there were four false-negatives. A total of 63 patients showed negative results out of 78 H.pylori-negative patients (specificity 80.76; 95% CI: 76.9-84.61%), and there were 15 false-positives. The positive predictive value of the stool antigen test was 67.3 (95% CI: 62.8-71.9%) and the negative predictive value of the stool antigen test was 94.03 (95% CI: 91.7-96.3%) [Table/Fig-6,7].
Performance of Epituub® fecal H.pylori antigen test kit.
Stool antigen test | H.pylori infection | Total | Chi-square | p-value |
---|
Present | Absent |
---|
1 | 0 |
---|
Positive | 31 (67.39%) | 15 (32.61%) | 46 | 48.12 | 0.001 |
Negative | 4 (5.97%) | 63 (94.03%) | 67 | | |
TOTAL | 35 | 78 | 113 | | |
Validation of Epituub® fecal H.pylori antigen test in relation to gold standard.
Validity | Percent | 95% CI | Chi-square | p-value |
---|
Sensitivity | 88.57 | 85.4-91.6 | 48.12 | 0.001 |
Specificity | 80.77 | 76.9-84.61 | | |
PPV | 67.39 | 62.8-71.9 | | |
NPV | 94.03 | 91.7-96.3 | | |
Accuracy | 83.1% | | | |
LR + | 4.585 | | | |
LR- | 0.142 | | | |
PPV: Positive predictive value; NPV: Negative predictive value; LR: Likelihood ratio
Discussion
The worldwide prevalence of H.pylori infection varies widely. A study conducted by Adlekha S et al., in Central Kerala detected a prevalence of 62% among dyspeptic patients (urease test and histopathology) whereas, Paul N et al., reported a prevalence of 36% from South Kerala (IgG antibody and urease test) [11,12]. Sodhi JS et al., detected 58% prevalence in Kashmir (RUT and histopathology) [13]. A study conducted by Rastogi M et al., showed a prevalence of 50.51% by SAT [14]. High prevalence of infection in present population justifies the use of ‘test and treat approach’ with non invasive tests [4]. Diagnostic tests with both high sensitivity and specificity, exceeding 90% are advisable for accurate diagnosis of H.pylori infection in clinical practice [15].
Differences in the antigenicity of H.pylori strains can affect the accuracy of SATs in different populations [16]. Also, the detection limit of bacterial antigen varies among different kits. Therefore, sensitivity and specificity of SATs should be tested in each population before use in the management of H.pylori infection [8]. Meta-analyses have shown that monoclonal antibody based assays are better compared to polyclonal antibody based assays [17]. Among the SATs ICT based tests are easy to perform and are useful for in-office rapid diagnosis of H.pylori infection [18]. In comparison to ELISA technique, ICT based tests do not require specialised equipment, technical expertise or a laboratory set up [19]. Validation of any test requires its comparison to a gold standard. In present study, a combination of two invasive methods, RUT and HPR are used as the diagnostic criteria for H.pylori infection which has served as the gold standard.
Poor socio-economic status and overcrowded conditions in developing countries have been attributed to the early childhood acquisition of H.pylori infection (30%-50%) which peaks during adulthood (over 90%) [20]. However, present study did not show any specific trend in the distribution of H.pylori infection among various age groups. Maximum H.pylori infection was in the age group 50-59 years (37%) and the number of patients with dyspepsia was also maximum in the same age group. There was no statistically significant difference in H.pylori infection above and below 45 years. Increasing trend of prevalence with age is not uniformly reported in all Indian studies. In a study of 500 adults, Ahmed KS et al., noticed increasing prevalence with age which peaked in the 70-79 year age group (90%; p<0.01) [21] while Khan S et al., noticed a maximum (74%) between 16-30 years and thereafter showing a decline [22].
In present study population, both dyspepsia and H.pylori infection were more in females (54% and 37% respectively) compared to males (46% and 23%) which were statistically non significant. The female preponderance in H.pylori positivity is contrary to other studies from India where male preponderance was reported without statistical significance [11,22]. It is well known that H.pylori infection is dependent upon many variables such as age, sex, socio-economic status, dietary habits, genetic, and immunological factors [3].
The test used in present study was Epituub® fecal H.pylori antigen Test kit which employs dye-conjugated monoclonal antibody against H.pylori antigen, and solid-phase/membrane coated specific anti-H.pylori monoclonal antibody. The detection limit of H.pylori is about 4-8 ng/mL. This rapid SAT had a sensitivity of 88.5% (95% CI: 85.4-91.6%), and specificity of 80.76% (95% CI: 76.9-84.610) in relation to gold standard. To the best of present knowledge, this is the first prospective study to examine the efficacy of the fecal H.pylori antigen test in the diagnosis of H.pylori infection from Kerala. The performance characteristics of different assays against similar gold standards have been published from across the world [23-26] [Table/Fig-8]. The sensitivity of kits range from 51% to 87% and specificity ranged from 87% to 95%.
Performance characteristics of different assays [9,23-26].
Authors (Year) | Country | Test | Patients | Gold standard | Sensitivity | Specificity |
---|
Rastogi M et al., [9] | India | Immunocard STAT HpSA test | 78 | HP, RUT | 95.5 | 81.1 |
Trevisani L et al., [23] | Italy | ImmunoCard STAT | 105 | RUT, HP | 85 | 93 |
Kesli R et al., [24] | Spain | H.pylori Fecal Antigen Test | 168 | RUT, HP | 81 | 92 |
Korkmaz H et al., [25] | Turkey | ImmunoCard STAT | 198 | RUT, HP | 68.9 | 92.6 |
One step H.pylori antigen | 198 | RUT, HP | 86.7 | 88.9 |
H.pylori Fecal Antigen Test | 198 | RUT, HP | 78.9 | 87 |
Korkmaz H et al., [26] | Turkey | Genx H.pylori CARD test | 162 | RUT, HP | 51.1 | 95 |
Present study | India | Epituub® fecal H.pylori antigen test | 113 | HP, RUT | 88.57 | 80.77 |
When compared to different rapid stool antigen methods, performance of present test kit was quite acceptable. A low rate of false negatives i.e., 11.5% may be due to factors like mild gastrointestinal bleed, poor bacterial colonisation of the stomach [25], possible use of bismuth containing antacids and a rare possibility of inadequate period of abstinence of PPI before SAT. Interference with other Helicobacter species may have caused false-positive test results [27,28].
The availability of validated non invasive SAT will facilitate the introduction of ‘test and treat strategy’ in the present study population. Easy and early detection of H.pylori infection and treatment may reduce the prevalence of associated complications such as chronic gastritis and carcinoma stomach in the long run.
Limitation and Future Recommendation
One major limitation of the study was the small sample size and hence larger studies are required to draw a conclusion regarding the utility of present method as a non invasive diagnostic test. A detailed analysis of sociodemographic factors is also required. Such studies will definitely help us to introduce the rapid SAT as an effective diagnostic or screening tool in present population.
Conclusion
In conclusion, this rapid ICT H.pylori SAT appears to have reasonable diagnostic accuracy in the pre-treatment setting, and could represent a valid alternative to the invasive tests. It is relatively easy, fast to perform and can reduce the endoscopic work load. Larger studies should be planned to confirm present results, and its use in the childhood population and in the post treatment setting. Large scale use of non invasive methods will make the tests cheaper and they can be introduced in primary care setting for early detection and treatment of H.pylori infection.
Financial support: The present study was conducted with the financial support from Institutional Research Committee, Government Medical College, Thrissur, Kerala, India with State Board of Medical Research (SBMR) grant.
Note: One patient was having both hiatus hernia and gastric ulcer. Another patient was having both proximal gastropathy and hiatus herniaPPV: Positive predictive value; NPV: Negative predictive value; LR: Likelihood ratio