Routine Anti-Helicobacter Immunohistochemical Staining is Significantly Superior to Reflex Staining Protocols for the Detection of Helicobacter in Gastric Biopsy Specimens

Background: The role of special stains in the detection of H. pylori is controversial; some pathologists claim that hematoxylin and eosin staining alone is adequate to detect bacteria, while others maintain that immunohistochemical staining must be used on all gastric biopsies. This study was designed to test the hypothesis that hematoxylin and eosin and histochemical special stains have similar sensitivity and that immunohistochemical staining only marginally enhances the detection rate of the organisms. Materials and Methods: Records of all gastric biopsies examined in a large pathology laboratory between January 1, 2010 and December 31, 2014 were extracted and stratified according to the stain used for the detection of Helicobacter (1. hematoxylin and eosin only; 2. Upfront HP Blue; 3. Upfront HP Blue including reflex immunohistochemistry; 4. Upfront anti-Helicobacter immunohistochemical staining. Results: Gastric biopsies from 794,859 endoscopies obtained from 622,945 unique patients were analyzed. The basic demographics of the patients in different staining groups were identical. A total of 56,955 gastric biopsy sets were stained with histochemical stains only, and 731,193 with anti-Helicobacter immunohistochemical stain; in 4409 of these, an immunohistochemical stain was also performed. In 2302 cases, only hematoxylin and eosin were performed. In this latter group, 7.0% of patients were Helicobacter-positive; in the upfront HP Blue, the prevalence was 7.8%, whereas patients with upfront immunohistochemistry had a prevalence of 10.2% (p <.0001, compared to hematoxylin and eosin only). Conclusions: There was a dramatic improvement in detection of H. pylori infection through the use of upfront immunohistochemical stains. Therefore, we contend that the routine use of IHC for all gastric biopsies is an appropriate protocol to ensure that every patient gets the right answer.