By Cory A. Roberts, M.D.
Helicobacter pylori is a common organism with widespread effects on the United States population. It is sometimes difficult to recall that this pathogen was first proposed as an associative agent with gastric cancer and peptic ulcer disease just over 20 years ago (1). We continue to learn about its role in these and other diseases and improve treatments. The accurate histologic detection of the organism remains a critical component of Gastrointestinal Pathology.
Helicobacter pylori (HP) is a microaerophilic fastidious gram negative organism. It is not easily cultured routinely. It requires an enriched transport medium if it is not placed on a culture plate within two hours of collection. It features a curved or comma shaped appearance. It can also change morphology after treatment which adds to the histologic detection problems. It can alter its form to a coccoid morphology which is non-culturable and loses its urease activity. The coccoid form is not likely to be recognized as HP on a biopsy specimen using either H&E or routine histochemical stains.
Helicobacter pylori is a class I (definite) human carcinogen as deemed by the World Health Organization and the International Agency for Research on Cancer (2). More than 90% of duodenal ulcers are associated with HP infection. It is a known cause of gastritis including atrophic gastritis, gastric ulcer, non-Hodgkin lymphoma (extranodal marginal zone lymphoma more commonly known as lymphoma of mucosa associated lymphoid tissue or MALT), and adenocarcinoma. It carries a 1.9 to 2.5 times the general population risk for adenocarcinoma (3,4). It is accepted that patients infected with HP ought to be treated and that the inherent ulcer risks associated with infection are not eradicated unless the organism is (5,6).
Initially, many pathologists simply sought out the organism via thorough inspection of the routine hematoxylin and eosin (H&E) stained slide. The organism can be seen on the H&E stain; however, careful scrutiny using a high power objective is often necessary and is likely to fail if organisms are few in number and will undoubtedly fail if the organisms are in the lamina propria (2). The H&E, while the least costly, simply lacks the sensitivity necessary to qualify as an adequate screening test.
A number of histochemical stains are currently employed in attempt to readily identify HP. Each has advantages and disadvantages. The first such stain is the Giemsa stain. The strength of the Giemsa stain lies in its ubiquity within histology laboratories. Even the smallest and least sophisticated laboratories can perform this stain with relative ease and reproducibility. Even undeveloped countries in Africa, which have a high HP prevalence and therefore concerns about false negatives, have shown relatively good sensitivity (85%), specificity (89%), positive predictive value (93%) and negative predictive value (74%) with the Giemsa stain (7). The stain is inexpensive, available and relatively easy to perform. It suffers from lack of specificity in that it will stain all bacteria and one must rely on classic HP morphology in interpretation. In addition, the stain must be examined on relatively high power in order to definitively identify the organism.
Silver based stains such as Warthin-Starry will also nicely mark the organism. These stains allow easier visualization of the organisms, particularly at lower power and may be more sensitive than the Giemsa. However, the silver based stains are less widely available. The reagents are more expensive and the stain is technically more difficult to perform. Therefore, experienced technologists who perform the stain more commonly are more important for the success of this stain than the simple Giemsa.
More recently, immunohistochemical (IHC) stains have become more commonplace. This type of staining relies on an antibody-antigen reaction. IHC for HP is a very sensitive (>90%) and specific test. It is easier to read for the pathologist and therefore easier to detect the organisms. One can detect fewer numbers of organisms relatively easily and even detect organisms in the lamina propria (2). Perhaps most importantly, IHC will still detect coccoid/post-treatment forms of HP which are not culturable or able to be accurately detected with the other stains. The drawbacks are that IHC is more technically difficult requiring strict internal controls and understanding and therefore less widely available. In addition, the stain is more expensive than Giemsa and silver based stains although the difference is negligible considering the cost of the whole procedure.
The Gastrointestinal Division of ProPath undertook a study to see how the Giemsa and IHC stains compared. The study proposal was to perform IHC on cases that were negative for HP on the initial Giemsa review yet had "suspicious inflammation" for HP and see if there was increased HP detection.
Thirty six biopsies on 35 patients were reviewed. Three of the 36 biopsies (8%) were positive for HP on initial Giemsa review and served as positive internal controls while the remaining cases (92%) were negative for HP on initial Giemsa review. Following IHC staining, 15 of 36 cases (42%) were positive for HP, an increase from the initial 8% positive rate (Figure 1). All 36 H&E and Giemsa stain cases were reviewed again (blinded) and a second review yielded four additional HP positive cases using only the Giemsa stain for a total of 7 out of 36 Giemsa stains positive for HP (19%). If one excludes the three positive internal controls (leaving 33 cases initially negative for HP) a second, close review of the Giemsa slide resulted in 4/33 positive (12%) while the IHC detected those four and an additional eight positive cases (12/33, 36%). The result was an increased detection of HP from 12% to 36% using IHC compared to Giemsa. There was no correlation between degree of inflammation (mild, moderate, marked) and the intensity of positive staining for HP.
Following this study, which supports other published works, we have now switched from routine performance of Giemsa stains on gastric biopsies to routine IHC. Our own Immunohistochemistry Division, headed by Dr. Rodney T. Miller, performs this and other stains with superb quality and same day service. We are convinced, after much thought and study, that the IHC is a better test for each of the reasons outlined above and will result in more patients receiving the all important antimicrobial treatment necessary to eradicate HP and to then confirm its absence on subsequent biopsies regardless of number organisms present or the shape it chooses.
1. Warren JR, Marshall BJ. Unidentified curved bacilli on gastric epithelium in active chronic gastritis. Lancet. 1983;1:1273-1275.
2. Jhala NC, Siegal GP, Klemm K, et al. Infiltration of Helicobacter pylori in the gastric mucosa. Am J Clin Pathol. 2003;119:101-107.
3. Weir S, Cuccherini B, Whitney AM, et al. Recurrent bacteremia caused by a "Flexispira"-like organism in a patient with X-linked (Bruton's) agammaglobulinemia. J Clin Microbiol. 1999;37:2439-2445.
4. Eslick GD, Lim LL, Byles JE, et al. Association of Helicobacter pylori infection with gastric carcinoma: a meta-analysis. Am J Gastroenterol. 1999;94:2373-2379.
5. NIH Consensus Conference. Helicobacter pylori in peptic ulcer disease. NIH Consensus Developmental Panel on Helicobacter pylori in Peptic Ulcer Disease. JAMA. 1994;272:65-69.
6. Neil GA. Do ulcers burn out or burn on? Managing duodenal ulcer diathesis in the Helicobacter pylori era. Ad Hoc Committee on FDA-Related Matters. Am J Gastroenterol. 1997;92:387-393.
7. Wabinga HR. Comparison of immunohistochemical and modified Giemsa stains for demonstration of Helicobacter pylori infection in an African population. Afr Health Sci. 2002;2:52-55.
Date of last revision: Fall 2004.