[Rapid detection of clarithromycin resistant Helicobacter pylori from children's gastric biopsy specimens by polymerase chain reaction-restriction fragment length polymorphism].

Objective: Infection with clarithromycin-resistant Helicobacter pylori (Hp) is often predictive of treatment failure. Susceptibility testing for Hp could guide therapy of Hp infections. However, agar dilution approved by the Clinical and Laboratory Standards Institute (CLSI) to test for antimicrobial susceptibility of Hp is time consuming (results are often not available in a week or more). So a more expeditious method is necessary. The purpose of this study was to evaluate polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) test performed directly on gastric biopsy specimen from children to detect 23S rRNA mutations (A2143G and A2144G) indicating clarithromycin resistance.

Methods: All biopsy specimens were derived from patients presenting with upper gastrointestinal symptoms, submitted to endoscopy in the Affiliated Children's Hospital, Zhejiang University School of Medicine from September 2006 to February 2007. No patients had undergone eradication therapy. Thirty-nine samples randomly selected from positive specimens by rapid urease test, were homogenized in 500 microl brucella broth with 30% glycerol. The 200 microl homogenized fluid was used to purify genomic DNA with the kit according to the instructions provided by manufacturer, and the rest was used to isolate Hp strains by culturing. All the Hp isolates were tested for clarithromycin susceptibility with the agar dilution and classified as resistant if the minimum inhibitory concentrations (MIC) exceeded 1 microg/ml. Simultaneously, PCR-RFLP analysis was performed in order to identify 23S rRNA mutations (A2143G and A2144G). Finally, the two methods were compared by statistics. The agar dilution was used as a standard to determine the sensitivity and specificity of the PCR-RFLP assay.

Results: Of the 39 samples, agar dilution and PCR-RFLP method respectively detected 13 (33.3%) and 14 (35.9%) clarithromycin-resistant gastric specimens. The sensitivity and specificity of PCR-RFLP for the detection of Hp in biopsy specimens were both 92%. The positive and negative predictive value was 85.7% and 96% respectively. No statistically significant difference was found between the two methods (chi2=0.06, P>0.05). The rate of Hp resistance to clarithromycin significantly increased compared with a previous report from the authors' hospital in 2004 (chi2=6.20, P<0.05).

Conclusions: Rising clarithromycin resistance rates were observed in children who visited the authors' hospital. PCR-RFLP test is reliable and rapid for detection of clarithromycin resistance directly on gastric biopsy specimen from children and may help choose appropriate antibiotic in Hp eradication therapy.