[13C]-urea breath test for assessment of cure of Helicobacter pylori infection at 1 month after treatment.

Although the [13C]-urea breath test has been reported to provide high specificity and sensitivity, its reliability in determination of cure, including when and how the assessment of cure is made, is not fully established. Here we evaluate the reliability of judgment of cure after 1 month of therapy using the [13C]-urea breath test, including the appropriate time to assess its cure. Our subjects were 199 H. pylori-infected patients (144 men and 55 women; 50.5+/-0.76 years) who underwent cure therapy, took the breath test at 1 month, and underwent both the breath test and invasive tests using endoscopy at 6 months after completion of therapy. Accuracy of the breath test in judgment of cure at 1 month was investigated by comparing the results at 1 month with those at 6 months when cut-off value of the breath test was set at 5 per thousand. Chronologic alterations of the delta13C value were also analyzed in cured patients who underwent the breath test at 1, 3, and 6 months. In 167 (94.9%) of 176 patients whose breath test values were less than 5 per thousand, cure was reconfirmed at 6 months. In 14 (60.9%) of 23 patients whose values were greater than 5 per thousand, cure was demonstrated at 6 months. The mean delta13C values of 33 cured patients who took the breath test at 1 month and repeated the breath test at 3 and 6 months for any reasons were 4.14+/-0.74 per thousand, 2.02+/-0.23 per thousand, and 1.72+/-0.19 per thousand, respectively, and the value at 3 months was significantly smaller than that at 1 month. The reliability of assessment of cure by only the breath test at 1 month was sufficiently high (94.9%) once the patients were judged to be cured. However, because the breath test at 1 month yielded a high false-positive rate and the breath test values at 3 and 6 months were similar, evaluation of cure of H. pylori infection by the [13C]-urea breath test should preferably be done at 3 months after treatment to minimize false-positive results, or the concept of a borderline group should be established for patients with positive but relatively high delta13C values (5-10 per thousand) who repeated the breath test 3 months after treatment.