Detection of Helicobacter pylori in virtual slides requires high resolution digitalisation

To diagnose Helicobacter pylori (HP) infection and its related mucosal injuries requires the histopathological analysis of gastric biopsies. The move from glass slides interpretation towards digital pathology implies technical choices to maintain the performances of histopathological diagnosis. The intra-rater agreement in assessing gastritis diagnostic criteria between glass slides, low resolution and high resolution digital slides in the subject of the present study. One hundred gastric biopsies were re-assessed by a single digestive pathologist on glass slides and digitalised slides at low resolution (ie, x20 magnification and single focus without z-stack) and high resolution (ie, x40 magnification with seven focus levels and z-stack) about the criteria of the updated Sydney system and the detection of HP. Inter-analyses agreement were very good (Kappa values>0.81) for every criteria but slightly inferior (ie, Kappa values<0.9) comparing glass slides interpretations with low resolution digital slides-based ones. Indeed, some HP infections were misdiagnosed using x20 magnification histochemical stained digitalised slides (p<0.05). At the opposite, anti-HP immunohistochemistry slides and/or x40 magnification digitalisation permitted to maintain almost perfect concordance in diagnosis (Kappa value>0.9). As mentioned in current guidelines, a high resolution x40 magnification digitalisation must be favoured in order to avoid some misdetection of microorganisms as HP.