Comparison of isotope ratio mass spectrometry and nondispersive isotope-selective infrared spectroscopy for C-13-urea breath test

OBJECTIVES: The C-13-urea breath test (UBT) is a sensitive and noninvasive method to diagnose Helicobacter pylori infection, but mass spectrometry (IR MS) is very expensive. The aims of this study were to compare the new low-p riced infrared spectroscopy with IRMS in detecting the infection and to ass ess the influence of feeding on test accuracy. METHODS: One hundred thirty- four patients with dyspeptic symptoms were recruited. Of these, 74 were inf ected and 60 uninfected on the basis of both CLO-test and histology. A subg roup of 37 patients (22 H. pylori-positive and 15 H. pylori-negative) was s tudied under fasting and nonfasting conditions on two different days. Dupli cate breath samples were analyzed with two IRMS systems (Breath Mat and ABC A) and an infrared spectrometer (IRIS) before, 15 min, and 30 min after ing estion of 75 mg C-13-urea with citric acid. In 37 patients the test was rep eated the day after the fasted one and was performed 60 min after a meal of 800 Kcal. RESULTS: There was a close correlation between IRIS and Breath M at (r = 0.969 at 15 min and r = 0.977 at 30 min; p < 0.0001), IRIS and ABCA (r = 0.963 at 15 min and r = 0.985 at 30 min; p < 0.0001), and Breath Mat and ABCA (r = 0.987 at 15 min and r = 0.981 at 30 min; p = 0.0001). The sen sitivity ranged from 97-100% at both times with all devices, although the s pecificity was slightly inferior with the infrared system than with the two IRMS machines (95% vs 98-100% at 30 min), but the difference was not signi ficant (p = NS). Food intake produced three false negative results in all t hree machines and a systematic shift to lower delta values in infected pati ents. CONCLUSIONS: Infrared spectroscopy can be considered a valid alternat ive to mass spectroscopy for the diagnosis of H. pylori infection. Fasting is required to guarantee an accurate test. (C) 1999 by Am. Cell. of Gastroe nterology.