Detection of Helicobacter pylori in adenotonsillar tissue of children with chronic adenotonsillitis using rapid urease test, PCR and blood serology: a prospective study.

Objective: Contradictory results have been reported regarding Helicobacter pylori (H. pylori) detection in adenotonsillar tissue. The aims of this study were to investigate whether adenotonsillar tissue of symptomatic children with chronic adenotonsillitis harbors the H. pylori organism, using two biopsy-based invasive methods namely; rapid urease test (RUT) and polymerase chain reaction (PCR) as well as blood serology and to compare the results obtained from each of these methods to the "gold standard".

Methods: This prospective clinical study was carried out on 20 children aged between 2 and 10 years scheduled for tonsillectomy +/- adenoidectomy in a tertiary referral center. Exclusion criteria included: use of antacids, H(2) blockers or antibiotics during the previous month before surgery and adenotonsillectomy for obstructive sleep apnea. Core biopsy samples from resected adenotonsillar tissue was tested for H. pylori detection using both RUT and PCR assay for the ureC gene. Preoperative patient venous blood samples were also tested for H. pylori IgG antibodies. As a "gold standard", examined tissue was considered to be H. pylori infected if the two biopsy specimen-based methods (RUT and PCR) yielded positive results.

Results: Thirty adenotonsillectomy specimens were tested (20 tonsils and 10 adenoids). RUT was positive in 16 (53.3%) specimens (12 tonsils and 4 adenoids). According to the "gold standard", 11/16 were considered false-positive, yielding this test sensitivity 100% and specificity 56%. The ureC gene sequence was detected by PCR in 5 (16.6%) specimens (3 tonsils and 2 adenoids), all of which were also positive by RUT, thus were considered H. pylori infected. Accordingly, PCR had a 100% sensitivity and specificity. Serology testing was positive for H. pylori IgG antibodies in 4/20 patients (20%), only two of them were found to have H. pylori infected adenotonsillar tissue.

Conclusions: Based on our findings it seems that adenotonsillar tissue may constitute an extra-gastric reservoir for H. pylori in symptomatic children with chronic adenotonsillitis. RUT was found to be of less accuracy than PCR in H. pylori detection in an extra-gastric location, thus results of previous studies using this test alone for detection of oral H. pylori should be treated with caution.