Introduction The exact incidence of Helicobacter pylori (HP) in childhood is unknown. There is a causal link between HP and development of peptic ulcers, gastric adenocarcinoma and lymphoma. However there is little evidence and guidance on the clinical management of HP in children. The North American Society for Paediatric Gastroenterology and Nutrition report there is no compelling evidence to treat children colonised with HP without histological evidence of gastritis. This reflects the current pathophysiological understanding of cancer development, where inflammation secondary to infection is believed to be the driver of dysplasia and cancer. In our tertiary referral centre current practice stains all paediatric gastric biopsies with Haematoxylin and Eosin stain (H+E) as well as Warthin Starry (WS) stain, specifically for HP-like organisms. The aim of this study was to investigate the incidence of HP infection in children undergoing endoscopy. We also examined the routine use of special staining on all gastric biopsies for HP and whether other histological features on H+E examination could act as reliable markers of infection allowing selective special staining.
Methods We reviewed all gastric biopsies submitted between January 8 and September 11, in patients born after 1989. All reports were loaded onto a database and searched for key words with “helico”, “HLO” and “H Pylori”. Each report was then examined for evidence of HP, acute inflammation and lymphoid aggregates.
Results 1102 cases were examined and 23/1102 were positive for HP, 2% of all cases. There was acute inflammation in 17/23 cases. In the six with no acute inflammation, two were suspected of having HP (one previously treated) and one had only a single biopsy from the body, not the antrum. 146/1102 (13%) cases had acute inflammation. Acute inflammation as a marker for HP has a sensitivity of 74%, specificity of 88%, PPV of 11% and NPV of 99%. Lymphoid aggregates and acute inflammation as markers of infection further increased HP detection.
Conclusion This study shows our current practice involves special staining of all biopsies for low prevalence disease (2%). However H+E examination alone has low sensitivity for picking up HP in cases of low density infection. Acute inflammation as a marker for HP infection could aid selection of cases for special staining. The NPV of 99% suggests that very few cases would be missed. Selective staining of cases for HP would also save around £3000/year (excluding pathologist time). The disadvantage that a small number of children colonised with HP, without gastritis may be missed is of unclear significance. Further research is required in order to develop an integrated clinical and pathological approach into the investigation, diagnosis and treatment of HP in children.
Competing interests None declared.