RESUMO
BACKGROUND: Numerous diagnostic tests are available to detect Helicobactor pylori (H. pylori). There has been no single test available to detect H. pylori infection reliably. We evaluated the accuracy of a new fluorescence quantitative PCR (fqPCR) for H. pylori detection in children. METHODS: Gastric biopsy specimens from 138 children with gastritis were sent for routine histology exam, rapid urease test (RUT) and fqPCR. 13C-urea breath test (13C-UBT) was carried out prior to endoscopic procedure. Gastric fluids and dental plaques were also collected for fqPCR analysis. RESULTS: 38 children (27.5%) were considered positive for H. pylori infection by gold standard (concordant positive results on 2 or more tests). The remaining 100 children (72.5%) were considered negative for H. pylori. Gastric mucosa fqPCR not only detected all 38 H. pylori positive patients but also detected 8 (8%) of the 100 gold standard-negative children or 11 (10.7%) of the 103 routine histology-negative samples. Therefore, gastric mucosa fqPCR identified 46 children (33.3%) with H. pylori infection, significantly higher than gold standard or routine histology (P<0.01). Both gastric fluid and dental plaque fqPCR only detected 32 (23.2%) and 30 (21.7%) children with H. pylori infection respectively and was significantly less sensitive than mucosa fqPCR (P<0.05) but was as sensitive as non-invasive UBT. CONCLUSIONS: Gastric mucosa fqPCR was more sensitive than routine histology, RUT, 13C-UBT alone or in combination to detect H. pylori infection in children with chronic gastritis. Either gastric fluid or dental plaque PCR is as reliable as 13C-UBT for H. pylori detection.