What diagnostic testing can be ordered to detect cytomegalovirus (CMV) colitis?

Diagnostic Testing for CMV Colitis

The gold standard for diagnosing CMV colitis is colonoscopy with tissue biopsy analyzed by CMV-specific immunohistochemistry (IHC) or tissue PCR, with blood serology having no diagnostic value. 1, 2

Primary Diagnostic Approach

Tissue-Based Testing (Gold Standard)

Colonoscopy with biopsy is mandatory when CMV colitis is suspected. The diagnostic algorithm should proceed as follows:

• Obtain adequate tissue samples: Minimum of 11 biopsies from the left colon in ulcerative colitis and 16 biopsies in Crohn's disease to achieve 80% detection probability 2, 3
• Target actively inflamed areas: Sample from ulcer bases and edges where CMV-positive cells are most concentrated 3
• Multiple colonic segments: Sample from multiple segments as CMV distribution varies by disease type 2

Send tissue for:

1. CMV immunohistochemistry (IHC) - This is the gold standard 1
2. Tissue PCR - Sensitivity 95%, specificity 66%; more sensitive than IHC but less specific 4
3. Hematoxylin & eosin (H&E) staining - Look for "owl eye" inclusion bodies, but this has poor sensitivity (10-87%) despite high specificity (92-100%) 5

Critical caveat: H&E alone misses most cases. The sensitivity of H&E is only 12.5% overall and 34.6% when compared to IHC 3. Always order IHC or tissue PCR in addition to routine histology.

Blood-Based Testing (Adjunctive Role)

Blood serology (IgG/IgM) has NO diagnostic value - 70% of adults are CMV seropositive, making serology useless for diagnosing active colitis 1

Serum CMV PCR has limited but specific utility:

• Sensitivity: 61-62%, Specificity: 90-91% 3, 6
• Best used as a negative predictor: A negative serum PCR effectively rules out CMV colitis 7
• Positive results (especially >250-259 IU/mL) support the diagnosis but cannot replace tissue confirmation 7, 8
• Does NOT correlate reliably with colonic infection 2

pp65 Antigenemia assay:

• Very low sensitivity (33-38%) but high specificity (94-98%) 3, 6
• Positive results confirm disease, but negative results don't exclude it
• Less useful than serum PCR 6

Stool-Based Testing (Emerging)

Stool CMV PCR shows promise but requires validation:

• Sensitivity: 53-70%, Specificity: 91-92% 4, 6
• May be useful when colonoscopy is contraindicated
• Not yet standard of care; further studies needed 4

Imaging

Contrast-enhanced CT scan is helpful but non-specific:

• Bowel wall thickening is almost always present 1
• Small bowel involvement (40% of CMV cases) helps differentiate from C. difficile 1
• Pancolic involvement is rare in CMV 1

Clinical Algorithm

For hospitalized IBD patients or those with moderate-to-severe refractory disease:

1. Order serum CMV PCR immediately - If negative, CMV colitis is unlikely 7
2. Proceed to colonoscopy with biopsy regardless of serum results if clinical suspicion remains high
3. Send tissue for both IHC and tissue PCR - Tissue PCR is most sensitive; IHC is most specific 5, 4
4. Do NOT order CMV serology - It provides no useful information 1
5. Consider stool CMV PCR if colonoscopy is contraindicated or high-risk 4, 6

Key Pitfalls to Avoid

• Never rely on H&E staining alone - It misses 65-88% of cases 3
• Never use blood serology for diagnosis - Useless due to high background seropositivity 1
• Don't assume negative serum PCR excludes disease - Proceed to tissue diagnosis if clinical suspicion is high, especially in severe colitis 2
• Avoid inadequate biopsy sampling - Too few biopsies dramatically reduce detection rates 2, 3
• Don't confuse low-level PCR positivity with clinically significant disease - Tissue viral load >250 copies/mg is the suggested threshold 3, though no universal cutoff exists 4

Special Considerations

In immunocompromised patients (transplant recipients, HIV, malignancy, purine analogue use):

• Lower threshold for testing 1, 2
• CMV colitis carries 70% mortality if untreated in severely ill patients 1
• Purine analogues are independent risk factors for CMV reactivation 2

Tissue PCR interpretation: Positive tissue PCR with high viral load (>250 copies) plus steroid refractoriness and systemic illness warrant treatment, even without classic histopathologic findings 2, 9. However, low-level tissue PCR positivity without clinical correlation may not require antiviral therapy 2.