Diagnostic Criteria for Eosinophilic Colitis
I must clarify that the evidence provided focuses almost entirely on eosinophilic esophagitis (EoE), not eosinophilic colitis (EoC), which is what you're asking about. The guidelines 1 all address esophageal disease, not colonic disease. However, I can provide diagnostic criteria for eosinophilic colitis based on the available research evidence.
Core Diagnostic Criteria
Eosinophilic colitis requires three essential components: (1) clinical symptoms of colonic dysfunction (abdominal pain, diarrhea, or constipation), (2) elevated eosinophil counts on colonic biopsy exceeding normal thresholds, and (3) exclusion of all secondary causes of colonic eosinophilia 2, 3.
Histological Thresholds
The histological diagnosis remains controversial due to lack of consensus, but the most commonly cited thresholds are:
- >40 eosinophils per high-power field (HPF) as a general cut-off that increases diagnostic specificity 2, 4
- Segment-specific thresholds (Collins criteria): >100 Eo/HPF in right colon, >84 Eo/HPF in transverse and left colon, >64 Eo/HPF in sigmoid and rectum 5
- >35 eosinophils per 100 epithelial cells in any segment 6
The threshold of >40 eosinophils/HPF (both mean and maximum counts) correlates better with clinically confirmed eosinophilic colitis and is associated with architectural distortion, intraepithelial eosinophils, and signs of eosinophil activation 4.
Supporting Histological Features
Beyond eosinophil counts alone, look for:
- Intraepithelial eosinophils
- Eosinophil microabscesses
- Signs of eosinophil degranulation
- Architectural distortion or mucosal atrophy
- Submucosal eosinophilic infiltration
- Absence of acute inflammation and lymphoplasmacytic infiltration (which suggests other diagnoses) 4
Clinical Presentation
The most common symptoms are:
- Chronic diarrhea (64% of cases) 5
- Recurrent abdominal pain (66% of cases) 5
- Less commonly: chronic constipation, GI bleeding, weight loss 5, 2
Key clinical clues that increase diagnostic likelihood:
- Peripheral blood eosinophilia (the main distinctive clinical characteristic) 6, 7
- Younger age and female gender 4
- History of atopy or allergic disorders (present in only 14% of cases) 5
Mandatory Exclusions (Diagnosis of Exclusion)
You must systematically rule out all secondary causes:
- Parasitic infections (stool studies, serologies)
- Drug-induced eosinophilia (NSAIDs, antibiotics, others)
- Food allergens (though skin testing has limited value in adults) 2
- Inflammatory bowel disease (IBD) - critical distinction 7, 8
- Graft-versus-host disease
- Connective tissue disorders
- Hypereosinophilic syndrome
- Malignancy
Endoscopic Findings
Critical pitfall: Colonoscopy is frequently normal or near-normal in eosinophilic colitis 6, 9. When abnormalities are present, they may include:
- Lymphoid nodular hypertrophy
- Nonspecific mucosal changes
- Findings that can mimic IBD 7
Therefore, systematic biopsies from multiple colonic segments are essential even with normal-appearing mucosa 6, 9.
Biopsy Protocol
Obtain biopsies from at least two different colonic segments, ideally including right colon, transverse colon, left colon, and rectum, as eosinophilic infiltration can be patchy 5, 3.
Diagnostic Algorithm
- Clinical suspicion: Patient with chronic diarrhea, abdominal pain, or constipation
- Check peripheral eosinophil count - if elevated, increases suspicion
- Colonoscopy with systematic biopsies from multiple segments (even if mucosa appears normal)
- Histological confirmation: >40 eosinophils/HPF with supporting features
- Exclude secondary causes: Comprehensive workup for parasites, medication review, rule out IBD
- Communicate findings: Proactive dialogue between endoscopist and pathologist is crucial to avoid misdiagnosis as IBD 7
Important Caveats
- Prevalence is 4.0-4.7% in patients with chronic diarrhea or IBS-D symptoms 6
- Histological persistence of eosinophils occurs in 75% of patients even with symptom relief, so don't rely on symptoms alone for monitoring 5
- Some patients may progress to IBD, particularly those with high eosinophil density and poor treatment response 8
- Higher prevalence in autism spectrum disorder (44% vs 28.5% in general population) warrants lower threshold for investigation in this population 5