Correspondence to: Dr B. C. Y. Wong, Department of Medicine, University of Hong Kong, Queen Mary Hospital, Pokfulam Road, Hong Kong, China. E-mail: firstname.lastname@example.org
Background : Recently, a rapid-release 100-mg 13C-urea tablet with citrate supplement (Diabact UBT) showed excellent performance in a European population.
Aim : To investigate the accuracy of a 50-mg tablet-based 13C-urea breath test protocol.
Methods : Consecutive dyspeptic patients referred for upper endoscopy were recruited. 13C-Urea breath test was performed using a 50-mg 13C-urea tablet (Diabact UBT) and compared with the gold standard (rapid urease test and histology). Baseline, 10-min, 20-min and 30-min breath samples were collected in all cases. The cut-off values at each measurement interval were determined by three standard deviations above the mean excess δ13CO2 excretion of Helicobacter pylori-negative patients.
Results : Two hundred patients (150 before therapy and 50 after therapy) were available for analysis, with a mean age of 48.4 years, and 99 patients (50%) were H. pylori positive. The sensitivity and specificity of the 50-mg tablet-based 13C-urea breath test at 10 min, 20 min and 30 min were 100% and 98%, 100% and 100%, and 100% and 98%, respectively.
Conclusion : A 20-min, 50-mg tablet-based 13C-urea breath test (Diabact UBT) protocol is highly accurate for the diagnosis of H. pylori infection.
The 13C-urea breath test (13C-UBT) is a technically simple and non-invasive test for the diagnosis of Helicobacter pylori infection. The accuracy of 13C-UBT in the diagnosis of H. pylori infection is high, with a sensitivity of 90–98% and a specificity of 92–100%.1–6 The sensitivity and specificity of post-treatment 13C-UBT are as high as those of pre-treatment 13C-UBT, making it a reliable test for the determination of post-treatment H. pylori status.7 Since the first description of 13C-UBT by Graham et al.,8 several modifications have been published to simplify and optimize the test, including the amount of 13C-urea used, type of test meal, number of samples and sampling time. Citrate has been described to be the best test meal.9,10 Previously, we have shown that a 50-mg 13C-UBT protocol with or without citrate, using a sampling point at 20 min, is an accurate and cost-saving approach for the diagnosis of H. pylori infection in the Chinese population.11 Recently, a newly developed rapid-release tablet containing 13C-urea and citric acid has been described.12 By enclosing the 13C-urea in a capsule, contamination by urease-producing oropharyngeal bacteria is avoided. This protocol has been investigated in a European population, with a sensitivity of 95% and a specificity of 100%.12 However, the dose of 13C-urea used was 100 mg (two tablets),12 which increased the cost of the test. On the basis of our previous experience with the 50-mg 13C-UBT protocol (powder), we postulate that this novel 13C-urea tablet at a dose of 50 mg may be equally reliable for the diagnosis of H. pylori infection.
Thus, the aims of this study were: (i) to investigate the 50-mg 13C-urea tablet protocol and determine the optimal cut-off value; (ii) to determine the optimal measurement interval of this protocol; and (iii) to investigate whether the accuracy of this protocol can be maintained in patients post-eradication.
Patients referred to the endoscopy unit of the Department of Medicine, Queen Mary Hospital, Hong Kong for the investigation of dyspepsia or for follow-up after H. pylori eradication and/or ulcer healing were recruited. Dyspepsia was defined as persistent or recurrent upper abdominal pain or discomfort over the preceding 3-month period. Informed written consent was obtained from all patients participating in the trial. Exclusion criteria included patients with previous gastric surgery and patients taking antibiotics, H2-receptor antagonists, bismuth compounds or proton pump inhibitors in the 4 weeks before endoscopy. Patients with a past history of H. pylori eradication were allowed to enter the study provided that the medication for eradication had been stopped for at least 4 weeks and the other inclusion and exclusion criteria were met.
During upper endoscopy, two antral biopsies and one corpus biopsy were taken. One antral biopsy was used for rapid urease test and the rest were sent for histological examination of H. pylori status by haematoxylin and eosin staining and modified Giemsa staining if necessary. For patients who had received eradication therapy within the past 3 months, an additional body biopsy was obtained for rapid urease test. The definition of H. pylori infection required both rapid urease test and histology to be positive and was used as the ‘gold standard’ in this study. Equivocal cases were excluded from the analysis. This approach has been validated previously in our centre, with a sensitivity and specificity of 100%, and less than 1% of cases cannot be diagnosed by this method.13
Each 13C-urea tablet (Diabact UBT, Diabact AB, Uppsala, Sweden) contained 50 mg of 13C-urea with 456 mg of anhydrous citric acid. The 50-mg tablet-based 13C-UBT was performed after an overnight fast. After the collection of a baseline exhaled breath sample in a vacutainer, the 50-mg 13C-urea tablet (Diabact UBT) was swallowed with 200 mL of water. Further breath samples were taken at 10, 20 and 30 min. No test meal was used before the breath test. All patients were recruited between December 2001 and May 2002. None of the recruited patients had participated in any previous 13C-UBT trials. Collected samples were analysed by a purpose-built isotope ratio mass spectrometer at the Simon K. Y. Lee Digestive Disease Laboratory, Queen Mary Hospital, Hong Kong. Results were expressed as the excess δ13CO2 excretion. The cut-off values at each measurement interval were determined by three standard deviations above the mean excess δ13CO2 excretion of H. pylori-negative patients. The sensitivities, specificities and accuracies of 13C-UBT were calculated at 10, 20 and 30 min.
The statistics used included Student's t-test, the chi-squared test and Fisher's exact test when appropriate. A P value of 0.05 or less was considered to be statistically significant.
Two hundred and nine patients were recruited. Nine cases were excluded from the analysis. Four cases had incomplete breath test data and five cases could not be classified according to our gold standard for the diagnosis of H. pylori infection, i.e. only one positive test out of the two tests (histology and rapid urease test). Thus, a total of 200 patients were available for analysis. Of these 200 patients, 50 had received eradication therapy within 3 months before the breath test, i.e. post-eradication patients. The mean age of the 200 patients was 48.4 years (range, 18–86 years). There were 87 males and 113 females. Ninety-nine per cent of the study patients were ethnic Chinese. Ninety-nine patients (50%) were diagnosed as H. pylori positive by the gold standard, 20 of which were post-eradication patients. Only one of the 20 H. pylori-positive post-eradication patients had a discordant rapid urease test between the antrum and the corpus (antral negative but corpus positive). Of these 99 H. pylori-positive patients, three had gastric ulcer (3%), 17 had duodenal ulcer (17%), one had gastric and duodenal ulcers (1%), 17 had gastro-duodenal erosions (17%) and one had oesophagitis (1%). One hundred and one patients (50%) were H. pylori negative by our gold standard. Of these 101 patients, two had duodenal ulcer (2%), one had gastric ulcer (1%), six had gastro-duodenal erosions (6%), one had gastric polyp (1%) and five had oesophagitis (5%). H. pylori status correlated strongly with the presence of duodenal ulcers (P < 0.001) and gastro-duodenal erosions (P = 0.006).
The calculated cut-off values were 1.2‰, 2.1‰ and 1.2‰ for the 10-min, 20-min and 30-min measurement points, respectively (Table 1). The sensitivity, specificity and accuracy of the 50-mg tablet-based 13C-UBT at 10, 20 and 30 min were 100%, 98% and 99%, 100%, 100% and 100%, and 100%, 98% and 99%, respectively. The 13C-urea tablet was easy to swallow and well tolerated by all patients. There was no difference in accuracy between patients before eradication therapy and those after eradication therapy (data not shown).
Table 1. Mean excess δ 13CO2 of Helicobacter pylori-negative patients (n = 101) at the 10-min, 20-min and 30-min sampling points
Mean excess δ13CO2
Cut-off (mean + 3 s.d.)
The mean excess δ13CO2 excretion at the 10-min (30.2 vs. 24.0, P < 0.001) and 20-min (29.8 vs. 24.0, P < 0.001) sampling points was significantly higher than that at the 30-min sampling point in H. pylori-positive patients before eradication therapy (Figure 1). Such a difference was also observed for the 20-min sampling point when compared with the 30-min sampling point in H. pylori-positive patients after eradication therapy (24.7 vs. 20.6, P = 0.004). The mean excess δ13CO2 excretion at the 20-min sampling point was similar to the reading at 10 min (29.8 vs. 30.2, P = 0.73) before H. pylori eradication. Similarly, the mean excess δ13CO2 excretion at the 10-min sampling point was similar to the reading at 20 min (25.4 vs. 24.7, P = 0.79) after H. pylori eradication. The reading at the 10-min sampling point was greater than that at 30 min (25.4 vs. 20.6, P = 0.14), after H. pylori eradication, but the P value was insignificant. However, the small number of patients (n = 20) in this particular subgroup may limit the power to detect a difference between the 10-min and 30-min measurements.
For H. pylori-negative patients, the mean excess δ13CO2 excretion was less than 0.3 at all three sampling points irrespective of eradication status (Table 1, Figure 1).
Five patients could not be classified according to our gold standard of H. pylori infection due to discordant results between rapid urease test and histology. Two patients had received eradication therapy previously. All five patients were invited to repeat a standard 13C-UBT using a validated 75-mg 13C-urea protocol.6 There was total agreement between the 50-mg tablet-based 13C-UBT and the conventional 75-mg 13C-UBT, suggesting that the discrepant results were due to the inaccuracy of the biopsy-based tests (Table 2).
Table 2. 13 C-Urea breath test ( 13 C-UBT) results of the five patients with discordant results between rapid urease test (RUT) and histology
50-mg tablet-based 13C-UBT
75-mg conventional 13C-UBT
13 C-UBT is one of the most important non-invasive methods for the detection of H. pylori infection. The low-dose 50-mg 13 C-UBT protocol with citrate as test meal has been validated in the Chinese population previously, with a sensitivity and specificity of 100% and 98%, respectively, using a cut-off of 3‰ and a measurement interval of 20 min. 11 The calculated cut-off values of the novel 50-mg Diabact UBT were 1.2‰, 2.1‰ and 1.2‰ for the 10-min, 20-min and 30-min measurement points, respectively. The sensitivity and specificity of the 50-mg Diabact UBT were 100% and 98%, 100% and 100%, and 100% and 98% at 10 min, 20 min and 30 min, respectively. The accuracy of this protocol is preserved in post-eradication patients.
At present, there is universal agreement that two samples should be taken: a baseline sample collected at time zero and a 30-min sample collected after the ingestion of 13C-urea. 13C is measured as the 13CO2 : 12CO2 isotope ratio and is expressed as the excess δ13CO2 excretion. The cut-offs in the three measurement intervals were determined by three standard deviations above the mean excess δ13CO2 excretion of H. pylori-negative patients. The cut-off determined by this method correlated well with that of Hamlet et al.12
The major advance of this novel 50-mg Diabact UBT is that it has an almost instantaneous disintegration and dissolution after entering the stomach.12 This rapid-release formulation is supported by the fact that the mean excess δ13CO2 excretion peaked at the 10-min and 20-min sampling points. For H. pylori-negative patients, the mean excess δ13CO2 excretion at the 10-min, 20-min and 30-min sampling points was close to zero, indicating that 13C-urea hydrolysis was minimal. Although this is not a head-to-head comparison, the mean excess δ13CO2 excretion values of this 50-mg tablet-based 13C-UBT at 10, 20 and 30 min were 0.16, 0.16 and 0.24, respectively, which are 5–10 times smaller than the corresponding mean excess δ13CO2 excretion values of 1.6 and 1.0 at 20 and 30 min (10-min measurement not available), respectively, using 50 mg of 13C-urea powder.11 Our data suggest that the 13C-urea tablet confers protection against oropharyngeal urease activity and discriminates well between H. pylori-positive and H. pylori-negative cases. Two H. pylori-negative patients showed false positive breath test results at the 10-min and 30-min sampling points according to our cut-off value. The corresponding excess δ13CO2 excretion values at 10 min, 20 min and 30 min in the first patient were 1.3, 1.3 and 1.8, respectively, and in the second patient were 2.1, 1.6 and 1.4, respectively. Although there was no statistical difference between the three time points, our data suggest that the 20-min sampling point has the highest accuracy. However, to power the study to detect a difference between the three time points would require an enormous investigation, which is probably inappropriate within the context of this study.
Five patients had discordant results between rapid urease test and histology; they were all invited for a conventional 75-mg 13C-UBT used in our laboratory.6 There was total agreement between the 50-mg tablet-based 13C-UBT and the conventional protocol, suggesting that the 50-mg tablet-based 13C-UBT may be more accurate than the biopsy-based tests.
The use of a test meal in 13C-UBT is employed to increase the residence time of 13C-urea in the stomach and to improve the contact between urease produced by H. pylori and the substrate. Citric acid has been suggested to be the best liquid test meal.9,10 The tablet formulation used in this study contains 456 mg of anhydrous citric acid. Although the dose of citrate is low when compared with that used in previous studies, the high accuracy obtained with the 50-mg 13C-urea tablet suggests that low-dose citrate is equally efficacious. This tablet is well tolerated, and the 13C-UBT can easily be performed outside the hospital/clinic, or even at home, after which a test-tube of expired air can be sent by regular mail for analysis. This may allow a potentially wider application of this non-invasive test for the diagnosis of H. pylori infection in regions at high risk of gastric cancer, such as China.
One of the limitations of this study is that we did not perform a head-to-head comparison of the 50-mg 13C-urea tablet with 50 mg of 13C-urea powder. As mentioned previously, in order to detect a small difference in accuracy between the tablet and powder, an enormous investigation would be required, which may be impractical.
In conclusion, we have shown that the 50-mg tablet-based 13C-UBT (Diabact UBT), using a measurement interval of 20 min, is highly accurate for the diagnosis of H. pylori infection.
This study was supported by the Peptic Ulcer Research Fund and the Simon K. Y. Lee Gastroenterology Research Fund, University of Hong Kong, Queen Mary Hospital, Hong Kong, China. We thank Swen Medico Ltd, Hong Kong, China for assisting the arrangement of the study. We would like to thank our research assistants F. M. Y. Fung, T. S. M. Tong and V. Y. K. Ho. We thank nurse specialist M. Chong and endoscopy nurses V. S. Y. Tang, D. K. K. Chang and W. P. Yung for their nursing assistance and provision of care to the patients.