[692] The Sydney System Twenty Years Later: Who Uses It and Does It Matter?

Jonathan G Lash, Robert M Genta. Caris Life Sciences, Irving, TX

Background: The original 1991 Sydney Gastritis Classification System recommended submitting two separate specimens each from antrum and corpus; the 1996 updated version added the incisura angularis (IA). The extent to which the Sydney System recommendations are followed remains unknown. This study was designed to determine the gastric sampling patterns in outpatient centers in the US and their diagnostic yield for Helicobacter infection and intestinal metaplasia (IM).
Design: We analyzed biopsy specimens submitted for the evaluation of gastritis from 400,738 endoscopic procedures performed between 1/1/08 and 6/30/11, using the Caris pathology database of biopsies received from outpatient endoscopy centers nationwide. Cases with an endoscopic impression of a discrete lesion were excluded. Based upon the requisition, biopsies were categorized as from: antrum, corpus, cardia, IA, or unspecified (NOS). The diagnostic yield for H. pylori infection (detected by IHC stain on all specimens) and IM were correlated with the number and sites of the specimens.
Results: There were 98,962 NOS biopsy sets; 264,604 specimens from the antrum; 69,737 from the corpus; 10,491 from the cardia; and 2,262 identified as IA. Specimens were from at least two sites (submitted in separate containers) in 33,072 procedures. Although neither version of the Sydney System was ever strictly followed, in 15,730 procedures (3.9%) there were at least two separate containers from antrum and corpus with at least two specimens each. The diagnostic yield for H. pylori and IM is summarized in Table 1.

Relative yield for different sets of gastric biopsies for the detection of Helicobacter and Intestinal Metaplasia

Biopsy set	Patients	H. pylori + (%)	OR (CI 95%)	IM (%)	OR (95% CI)
≤3 (A or C)	225,927	19,862 (8.8)	1	7,163 (3.2)	1
≥4 (A or C)	60,394	7,110 (11.8)	1.38 (1.35 - 1.42)	2,941 (4.9)	1.56 (1.50 - 1.63)
≥2A and ≥2C	15,730	2,324 (14.8)	1.80 (1.72 - 1.88)	945 (6.0)	1.95 (1.82 - 2.09)

A = Antrum; C = Corpus (2 specimens designated as antrum and 2 as corpus are compliant with the Sydney System.)


Conclusions: The Sydney System recommendations are rarely followed in the US. Compared to 3 or fewer biopsies, the diagnostic yield for both H. pylori and IM is nearly doubled by the submission of at least 2 separate antral and 2 corpus specimens. Cardia biopsies have the lowest yield. The detection of IM increases with the number of biopsies from any site (but cardia). The IA, originally added for its suspected early development of IM, yields only 3.5%. Sampling at least 2 antral and 2 corpus specimens is the best-yielding practice for the evaluation of gastritis.
Category: Gastrointestinal

Monday, March 19, 2012 1:00 PM

Poster Session II # 99, Monday Afternoon