MALDI Imaging zur Untersuchung der Helicobacter pyloriassoziierten Gastritis

Aims: Current diagnosis of Helicobacter Pylori gastritis relies on the individual histopathological examination of the gastric biopsy specimen. The aim of our study is to establish imaging mass spectrometry (IMS) technology as a new screening tool for gastric disease associated with Helicobacter pylori infection, to support pathological diagnosis. An alternative working flow for diagnosing gastritis is proposed. Methods: We investigated 55 infected and 26 non-infected bioptic samples of gastric mucosa of the antrum and corpus regions. FFPE samples were deparaffinized, antigen retrieved and on-tissue digested with trypsin. Subsequently, MALDI analysis was performed using an Autoflex Speed mass spectrometer. Mass spectra from regions of interest were exported for classification analysis. Four sample groups, antrum infected (cases=28), corpus infected (cases=27), antrum normal (cases=13), and corpus normal (cases=13), were created to produce an average spectrum representative of each group. Mass spectra were exported and loaded into ClinProTools software for classification analysis. Results: We first investigated whether it was possible to detect specific signals discriminating infected and normal samples. Four m/z values (896, 940, 1213, 1836) showed differential expression in the Helicobacter pylori infected samples. We then investigated whether it was possible to detect specific signals correlated to bacterial peptides in the antrum comparing infected with non-infected antrum tissues to build a classification model using a Support Vector Machine (SVM) algorithm. SVM incorporated eight ion peptides (896, 1213, 1615, 1629, 1693 1836, 3100, 3101) through statistical comparison (Wilcoxon/Kruskal-Wallis test). The class-prediction model generated was then applied to a test set of nine infected samples which could classify the gastritis subgroup with an accuracy of 93%. Conclusion: IMS can be used to detect Helicobacter pylori in bioptic gastric samples to differentiate infected from non-infected tissues. Single mass-to-charge ratios can be attributed to specific proteins to identify the protein structure of the analyzed tissues. This methodology is able to analyze multiple samples simultaneously. Cost and time effective diagnosis of B-gastritis is possible by MALDI mass spectrometry. Histopathological investigation may be the second step of the diagnostic procedure after exclusion of bacterial infection by MALDI mass spectrometry.