The Sydney Protocol is a standardized method of collecting five biopsies from specific, mapped locations in the stomach so that pathologists can reliably assess inflammation, atrophy, intestinal metaplasia, and Helicobacter pylori infection across the entire organ. Rather than taking a single sample from wherever looks abnormal, the protocol treats the stomach as a patchwork of functionally different zones and samples each one, catching precancerous changes that a targeted biopsy from just one spot would miss. Despite being the international standard for over two decades, real-world compliance remains strikingly low, and the gap between what guidelines recommend and what actually happens during endoscopy has meaningful consequences for patients at risk of gastric cancer.
Where the Five Biopsies Come From
The Updated Sydney System biopsy protocol (often shortened to “Sydney Protocol” or USSBP) calls for tissue samples from five defined sites. Two come from the antrum, the lower portion of the stomach near its outlet: one from the lesser curvature and one from the greater curvature, both within two to three centimeters of the pylorus. Two more come from the corpus, or body of the stomach: one from the lesser curvature roughly four centimeters above the angle, and one from the greater curvature about eight centimeters below the top of the stomach. The fifth biopsy comes from the incisura angularis, the sharp bend where the body meets the antrum.1PubMed Central. Implementation of the Updated Sydney System biopsy protocol improves the diagnostic yield of gastric preneoplastic conditions: results from a real-world study – Section: Gastric biopsies sampling method.
Each specimen is placed in a separately labeled container so the pathologist knows exactly which part of the stomach it came from. This matters because disease does not spread evenly. Atrophy might be advanced in the antrum but absent in the corpus, or vice versa. Without site labeling, the pathologist sees inflammation but cannot say where it lives, which makes risk assessment nearly impossible.
The protocol also provides a visual analogue scale for grading five histological features: mononuclear cell infiltration, neutrophil activity, glandular atrophy, intestinal metaplasia, and H. pylori density. These scales give pathologists a reference image for each severity grade, reducing the subjectivity of reading slides.2PubMed. Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston 1994
Why Five Sites Instead of One or Two
The whole point of mapped biopsies is that precancerous changes in the stomach are patchy. A single biopsy from one location can easily land on normal-looking tissue while atrophy or intestinal metaplasia hides a few centimeters away. Studies comparing the Sydney Protocol against single-site or fewer-site approaches consistently show a large diagnostic gap.
A cross-sectional study comparing 50 patients who received Sydney Protocol biopsies against 50 who received targeted single-site biopsies illustrates the difference starkly. The protocol-compliant group had chronic atrophic gastritis detected in 60% of cases, compared to 28% in the single-site group. For intestinal metaplasia, the numbers were 20% versus 8%. And the protocol group uncovered three patients with stage III atrophic gastritis and one with stage IV intestinal metaplasia, stages that carry real cancer risk and that were invisible in the single-site arm.3JOURNAL OF CLINICAL AND DIAGNOSTIC RESEARCH. Precision versus Protocol: A Cross-sectional Analysis of Gastric Biopsy Techniques in Detecting Premalignant Lesions – Section: Results
Another study looked at what happens when you take only one or two biopsies from the antral lesser curvature, a common shortcut. If a single biopsy from that location was the only sample, more than half the patients who actually needed follow-up surveillance would have been missed. Even adding a second antral biopsy still missed about a third of them.4PubMed Central. Comparison of the diagnostic accuracy of the updated Sydney system and single biopsy The incisura angularis biopsy is particularly valuable because that bend is often the first site to develop atrophy and metaplasia, and skipping it can leave dangerous changes undetected.
How Sydney Biopsies Feed Into Cancer Risk Staging
Collecting mapped biopsies is not just about finding inflammation. The location-specific samples are the raw material for two staging systems, OLGA and OLGIM, that estimate a patient’s risk of progressing to gastric cancer. OLGA stages atrophy by combining the severity of glandular loss across the antrum (including the incisura) and the corpus. OLGIM does the same but focuses exclusively on intestinal metaplasia, a change where stomach lining starts to resemble intestinal tissue. Both systems produce a stage from 0 to IV, with stages III and IV considered high risk.
A long-term follow-up study of over 7,400 patients used mapped gastric biopsies and OLGA staging to track who developed cancer over time.5PubMed. OLGA Gastritis Staging for the Prediction of Gastric Cancer Risk: A Long-term Follow-up Study of 7436 Patients Without the site-specific biopsy data, those stages could not have been calculated, and the patients at highest risk could not have been identified for surveillance.
These staging systems only work, though, if the biopsies actually come from the right places and get submitted in separately labeled jars. When a clinician sends unlabeled tissue to the lab, the pathologist can grade inflammation and metaplasia but cannot assign an OLGA or OLGIM stage because there is no way to tell antral tissue from corpus tissue under the microscope. The staging system becomes useless.
The Interobserver Agreement Problem
Even with properly mapped biopsies, different pathologists sometimes disagree on what they see. This has been a persistent concern with OLGA staging in particular, because atrophy can be subtle and the line between “mild” and “moderate” is judgment-dependent. European guidelines have noted the inconsistency in histopathological diagnosis and severity assessment between pathologists when staging gastric atrophy.6Gut. British Society of Gastroenterology guidelines on the diagnosis and management of patients at risk of gastric adenocarcinoma – Section: Gastric atrophy
A study comparing expert gastrointestinal pathologists found moderate agreement for atrophy across the three biopsy zones (antrum, incisura, and corpus), but much stronger agreement for intestinal metaplasia, which is easier to identify under the microscope because the morphological changes are more distinctive. General pathologists without subspecialty training performed worse, with poor agreement on atrophy. This is partly why OLGIM, which relies on metaplasia rather than atrophy, tends to show better interobserver reproducibility.7PubMed. Gastritis staging: interobserver agreement by applying OLGA and OLGIM systems
However, using OLGIM alone creates its own problem. The same study found that a substantial proportion of patients at high risk by OLGA criteria would be missed if staging relied only on OLGIM. The researchers recommended using both systems together to balance reproducibility against sensitivity. More recent data from a cross-sectional study reported excellent agreement for both systems, with kappa values above 0.9, though the study design and pathologist training likely influenced the results.8JOURNAL OF CLINICAL AND DIAGNOSTIC RESEARCH. Interobserver Agreement in Histopathological Assessment of the Operative Link on Gastritis Assessment and the Operative Link on Gastric Intestinal Metaplasia Staging: A Cross-sectional Study – Section: Results The takeaway is that interobserver variability is real but improving, and combining OLGA and OLGIM mitigates the weaknesses of each.
Real-World Adherence Is Alarmingly Low
Here is where the gap between guideline and practice is widest. A massive analysis of over 400,000 sets of gastric biopsies found that only 3.9% were submitted in a way that complied with the Sydney System, meaning separate containers with at least two antral and two corpus specimens. Two-thirds of all biopsy sets were labeled simply as “antrum” with no corpus sample at all, and nearly a quarter arrived at the lab without any topographic identifier whatsoever.9PubMed. Adherence to the Sydney System guidelines increases the detection of Helicobacter gastritis and intestinal metaplasia in 400738 sets of gastric biopsies – Section: RESULTS
This is not a trivial shortcut. The same study showed that Sydney-compliant sets detected H. pylori gastritis and intestinal metaplasia at substantially higher rates than non-compliant ones. A separate real-world study found that institutions that regularly followed the protocol detected chronic atrophic gastritis at roughly four times the rate of institutions that used it infrequently, and autoimmune gastritis at more than six times the rate.1PubMed Central. Implementation of the Updated Sydney System biopsy protocol improves the diagnostic yield of gastric preneoplastic conditions: results from a real-world study – Section: Gastric biopsies sampling method.
Why is adherence so poor? Multiple factors converge. Taking five biopsies from mapped locations adds time to the endoscopy, particularly when the endoscopist needs to label separate containers. In busy clinics, the pressure to move quickly is intense. Many endoscopists were trained before the protocol became standard, or trained in settings where gastric cancer risk is perceived as low. And in some health systems, the additional biopsy processing costs are not clearly reimbursed, discouraging thoroughness. The result is a protocol that everyone agrees works but that most practices do not actually follow.
How Advanced Imaging Changes the Equation
Narrow-band imaging (NBI) and other enhanced visualization technologies allow endoscopists to see mucosal surface patterns and vascular architecture in real time, highlighting areas of metaplasia that look normal under regular white light. This raises an obvious question: if you can see the lesion, do you still need random mapped biopsies?
The evidence says you need both. A study of 95 patients compared NBI-targeted biopsies alone against NBI-targeted biopsies combined with Sydney Protocol random biopsies. Targeted biopsies by themselves achieved about 88% sensitivity and 90% specificity for detecting intestinal metaplasia. But combining them with the Sydney Protocol pushed sensitivity to 100% and accuracy to nearly 95%. On their own, the targeted approach missed extensive metaplasia in about 5% of patients.10PubMed Central. Performance status of targeted biopsy alone versus Sydney protocol by non-NBI expert gastroenterologist in gastric intestinal metaplasia diagnosis – Section: Results
That 5% miss rate matters because the patients being missed tend to be the ones with widespread disease, exactly the group most at risk for cancer. NBI is excellent for directing attention to the most suspicious-looking areas, but it does not replace systematic sampling from all five mapped sites. The current European guidelines recommend using both approaches together: high-quality endoscopy with image enhancement to inspect the mucosa, plus protocol biopsies to catch what the eye cannot see.11PubMed. Screening for gastric premalignant lesions with narrow band imaging, white light and updated Sydney protocol or both? – Section: METHODS
Surveillance Intervals Based on Biopsy Results
Once mapped biopsies have been collected and staged, the results determine whether a patient needs ongoing monitoring and how often. The most recent European guidelines, updated in 2025 by a consortium of gastroenterology and pathology societies, split patients into two broad categories. Those with advanced stages of atrophic gastritis, meaning severe atrophic changes or intestinal metaplasia involving both the antrum and corpus (OLGA/OLGIM stage III or IV), should have a high-quality repeat endoscopy every three years regardless of where they live.12PubMed. Management of epithelial precancerous conditions and early neoplasia of the stomach (MAPS III): European Society of Gastrointestinal Endoscopy (ESGE), European Helicobacter and Microbiota Study Group (EHMSG) and European Society of Pathology (ESP) Guideline update 2025
For patients with mild to moderate atrophy confined to the antrum and no other risk factors, routine surveillance is not recommended. The earlier 2019 version of these guidelines added a nuance: if a patient has metaplasia at a single location but also has a family history of gastric cancer, or incomplete-type intestinal metaplasia, or ongoing H. pylori infection, then surveillance every three years may still be warranted.13PubMed. Management of epithelial precancerous conditions and lesions in the stomach (MAPS II): European Society of Gastrointestinal Endoscopy (ESGE), European Helicobacter and Microbiota Study Group (EHMSG), European Society of Pathology (ESP), and Sociedade Portuguesa de Endoscopia Digestiva (SPED) guideline update 2019
None of these decisions can be made without site-mapped biopsies. A pathology report that says “chronic gastritis with intestinal metaplasia” but does not specify whether the metaplasia is in the antrum alone or also the corpus leaves the clinician guessing. The difference between “antrum only, mild” and “antrum plus corpus, moderate” is the difference between no surveillance and a follow-up endoscopy every three years, potentially for the rest of that patient’s life.
Serum Pepsinogen as a Non-Invasive Screening Alternative
Taking five biopsies requires an endoscopy, which is invasive, expensive, and not available everywhere. Serum pepsinogen testing offers a blood-based alternative for screening populations before endoscopy. Pepsinogen I is produced mainly in the corpus, and when corpus glands atrophy, levels drop. The ratio of pepsinogen I to pepsinogen II also falls with advancing atrophy. In theory, low pepsinogen levels could flag patients who need endoscopy with protocol biopsies while sparing those with healthy stomachs.
In practice, the test has meaningful limitations. A Korean study evaluating pepsinogen I levels at a cutoff of 100 ng/mL found sensitivity of 75% to 86% for detecting histological atrophy and metaplasia, but specificity was only 32% to 35%. Using the pepsinogen I/II ratio at a cutoff of 5.3 improved specificity to the 70%–74% range but dropped sensitivity to 50%–69%.14Gut and Liver. Role of Serum Pepsinogen Tests in Detection of Gastric Atrophy, Intestinal Metaplasia, Gastric Adenoma, and Gastric Cancer in South Korea – Section: Results
What these numbers mean in plain terms is that the blood test catches most people with atrophy but also flags a lot of people who turn out to be fine, and depending on which threshold you use, it can miss a sizable minority who actually do have precancerous changes. Pepsinogen testing is best understood as a triage tool for large populations, not a replacement for biopsy-based staging. It can identify who should get a thorough endoscopy with mapped biopsies, but it cannot replace the information those biopsies provide about the location and severity of disease.
Artificial Intelligence for Grading Sydney System Biopsies
One of the most promising developments for improving the consistency of biopsy interpretation is the use of AI to grade the five Sydney System features. A recently validated system called SydneyMTL, trained on over 50,000 whole-slide images from routine clinical practice, jointly predicts all five attributes: mononuclear cell infiltration, neutrophil activity, glandular atrophy, intestinal metaplasia, and H. pylori density. Against the readings of 24 board-certified pathologists, the system achieved a mean agreement rate of about 89%, with over 80% agreement for 21 of the 24 readers.15PubMed. Comprehensive and reproducible grading of the Updated Sydney System in gastric biopsies using artificial intelligence: multi-pathologist validation and a reader study
In a randomized crossover study, pathologists who used the AI assistance showed improved interobserver agreement and reduced their review time by about a third for complete Sydney grading. This directly addresses the reproducibility gap described earlier. If AI can provide a standardized first-pass reading of all five histological features, the pathologist’s role shifts from primary grading to quality review and adjudication of difficult cases. For busy community pathology labs where subspecialty GI expertise may be limited, that kind of decision support could narrow the diagnostic gap between expert and general pathologists.
The technology does not solve the upstream problem of clinicians not taking mapped biopsies in the first place. AI can only grade what it receives. But if the bottleneck on the pathology side, namely inconsistent grading and the time required for thorough assessment, becomes less of a barrier, it removes one more excuse for skipping the protocol. A system that automates grading and delivers a provisional OLGA/OLGIM stage within minutes of slide scanning could eventually be integrated into the endoscopy suite workflow, giving real-time risk stratification before the patient even leaves the procedure room.
When the Protocol Might Not Be Enough
Five biopsies from standardized locations represent the minimum for comprehensive gastric mapping, but some clinical situations call for more. Patients with a strong family history of gastric cancer, those with known extensive metaplasia, or individuals undergoing evaluation for autoimmune gastritis may benefit from additional biopsies beyond the five standard sites. Some clinicians routinely take seven or more samples, adding extra corpus biopsies in suspected autoimmune cases because the disease disproportionately affects the oxyntic mucosa of the body and fundus.
Any visible lesion, whether a polyp, ulcer, or area of discoloration, should be biopsied separately and submitted in its own container regardless of the standard protocol sites. The Sydney Protocol maps background mucosal disease; it was never designed to replace targeted sampling of focal abnormalities. The two approaches are complementary. One tells you about the general state of the stomach lining, and the other tells you what a specific suspicious lesion is.
There are also populations where the standard five-site approach has received less validation. Most of the foundational research on the Sydney System was conducted in European and East Asian populations with relatively high rates of H. pylori infection and gastric cancer. In populations where H. pylori prevalence is low and gastric cancer is rare, the clinical utility of routine mapped biopsies during every upper endoscopy is debated. American guidelines, for example, have been slower than European ones to mandate the full protocol during routine procedures, partly reflecting different baseline risk profiles and partly reflecting cost considerations in a fee-for-service health system. The ongoing tension between thoroughness and efficiency is unlikely to resolve uniformly across healthcare settings.