Recommended Biopsy Sites for Eosinophilic Esophagitis
Take at least six biopsies from a minimum of two different esophageal levels—specifically the distal and proximal esophagus—combining both targeted biopsies from visibly abnormal mucosa and non-targeted biopsies from normal-appearing areas. 1
Optimal Biopsy Protocol
The most recent and highest quality guidelines from the British Society of Gastroenterology (2022) provide the strongest evidence for biopsy strategy:
Number and Location
- Six total biopsies minimum from at least two anatomical sites 1
- Primary sites: Distal AND proximal esophagus 1
- This approach achieves 97-100% diagnostic sensitivity 1
Targeting Strategy
Combine two approaches in your six biopsies:
- Targeted biopsies: Sample visible abnormalities (white spots/exudates, furrows, edema, rings) 1
- Non-targeted biopsies: Sample normal-appearing mucosa 1
This dual approach is critical because eosinophilic infiltrate distribution is patchy, and up to 10% of patients have isolated non-distal disease that would be missed with distal-only sampling 2.
Why This Matters
The Patchy Nature Problem
The eosinophilic infiltrate in EoE is notoriously patchy 1. Recent 2025 data shows that 56% of newly diagnosed patients had at least one esophageal segment with <15 eosinophils/hpf, and 10% had exclusively non-distal disease 2. Missing the middle esophagus would miss approximately 2% of active disease cases 3.
Distal vs. Mid vs. Proximal Distribution
While the 2018 international consensus recommended "two or more esophageal levels" without specifying exact sites 4, newer evidence clarifies the hierarchy:
- Distal esophagus: Highest yield—eosinophil density tends to be increased distally 5, 6
- Mid esophagus: Critical addition—isolated mid-esophageal disease occurs in up to 10% of cases 2
- Proximal esophagus: Important but rarely shows isolated disease (no cases of isolated proximal eosinophilia in recent studies) 2
When mid and proximal biopsies are discordant, 94% show mid-esophageal eosinophilia rather than proximal 2, supporting the distal + mid combination as optimal.
Clinical Pearls
Endoscopic Findings Guide Yield
Biopsies from areas with exudates and furrows demonstrate significantly higher eosinophil counts 5, 6. Severe exudates and lower esophagus location are independent predictors of positive findings 5. However, normal-appearing mucosa can still harbor diagnostic eosinophilia, so non-targeted biopsies remain essential 1, 6.
Age and Gender Considerations
Increasing age and male gender independently predict non-distal disease (OR 1.02 and 1.89 respectively) 2, making comprehensive sampling even more important in older male patients.
What About Gastric/Duodenal Biopsies?
Routine gastric and duodenal biopsies are NOT recommended unless there are specific symptoms or endoscopic abnormalities 4. Recent 2026 data shows only 20.9% diagnostic yield for extra-esophageal findings, predominantly H. pylori gastritis (18.8%), with clinically meaningful diagnoses like celiac disease occurring in only 0.6% 7. The 2018 consensus states these should be obtained "as clinically indicated" but have low yield in asymptomatic patients 4.
Common Pitfalls to Avoid
- Taking only distal biopsies: Misses 10% of cases with non-distal disease 2
- Sampling only abnormal-appearing mucosa: The patchy distribution means normal mucosa may harbor diagnostic eosinophilia 1, 6
- Taking fewer than 4-6 biopsies: Probability of diagnosis drops significantly below four fragments 8
- Assuming rings alone indicate active disease: Rings without furrows/plaques may persist in treated EoE and don't reliably correlate with elevated eosinophil counts 6
Practical Algorithm
For initial diagnosis:
- Obtain 6 biopsies total
- Sample distal esophagus (3 biopsies)
- Sample proximal OR mid esophagus (3 biopsies)
- Include both visibly abnormal areas AND normal-appearing mucosa
- Skip gastric/duodenal biopsies unless symptomatic or endoscopically abnormal
For follow-up monitoring: The same protocol applies, though recent 2026 data suggests two-site sampling (distal + proximal) maintains 98% accuracy with only 2% miss rate 3, potentially allowing streamlined follow-up protocols.