What are the CHOI criteria for response assessment in gastrointestinal stromal tumors (GISTs) and what are their clinical implications compared to RECIST, especially for tyrosine‑kinase inhibitor therapy such as imatinib?

CHOI Criteria for Response Assessment in GISTs

In GISTs treated with tyrosine kinase inhibitors like imatinib, response assessment should incorporate both tumor size AND tumor density on CT scan, as recommended by ESMO-EURACAN guidelines, because conventional size-based RECIST criteria frequently miss early responses and can misclassify responding tumors as stable or even progressive disease 1.

Why CHOI Criteria Matter

The fundamental problem with RECIST in GIST is that imatinib-responding tumors often show decreased tumor density and increased homogeneity on CT before any size reduction occurs—or even while tumors are increasing in size 1. This creates a critical clinical pitfall: a tumor that appears "stable" or "progressive" by RECIST may actually be responding well to therapy.

Key Response Patterns to Recognize

Responding GISTs demonstrate:

• Decreased tumor density on CT (becoming more hypodense/cystic)
• Increased homogeneity
• Reduced vascularity
• These changes may precede or occur without tumor shrinkage 1

A tumor can even increase in size yet still represent treatment response if density decreases 1. This counterintuitive finding is why size-only criteria fail in GIST.

Clinical Implications

1. Avoid Premature Treatment Changes

The most dangerous pitfall is discontinuing effective imatinib because RECIST shows "progression" when the tumor is actually responding. Always confirm apparent early progression with an experienced team before changing therapy 1. Consider:

• FDG-PET scan for early response assessment (highly sensitive, shows metabolic response within weeks) 1
• Reassessment of CT density changes
• Ruling out "false progression" patterns

2. Recognize "Nodule Within a Mass"

True progression often manifests as increased density within previously responding lesions—the classic "nodule within a mass" pattern where a hyperdense nodule appears within a hypodense responding tumor 1. This represents focal resistance and may warrant focal treatment (surgery, ablation) while continuing imatinib 2.

3. New Lesions May Not Be Progression

The "appearance" of new lesions can result from easier detection of less dense tumors that were previously invisible 1. Don't automatically classify this as progression without considering density changes.

4. Absence of Progression = Response

In GIST, stable disease at 6 months is considered treatment response 1. This differs from most solid tumors and reflects the unique biology of TKI-treated GIST.

Practical Response Assessment Algorithm

At each follow-up CT, evaluate:

1. Tumor density (Hounsfield units on contrast-enhanced CT)

   • Decreased density = likely response
   • Increased density in previously responding areas = likely progression

2. Tumor size (unidimensional or bidimensional)

   • May decrease, stay stable, or even increase with response

3. Tumor homogeneity

   • More homogeneous/cystic = response
   • New solid/hyperdense components = progression

4. Consider FDG-PET when:

   • Doubt exists about response status
   • Early prediction needed (e.g., preoperative assessment)
   • Note: Small proportion of GISTs lack FDG uptake 1

Evidence Quality Considerations

While the ESMO-EURACAN guidelines strongly recommend incorporating density criteria 2, 1, research evidence on Choi criteria specifically shows mixed predictive value. Studies demonstrate that while Choi criteria detect more responses than RECIST, the distinction between partial response and stable disease by Choi criteria may not predict different progression-free or overall survival 3, 4. What matters most is distinguishing responders (PR or SD) from true progressors (PD)—both RECIST and Choi accomplish this similarly 3.

The 2022 ESMO guidelines emphasize that both size and density should be considered together rather than rigidly applying any single criteria system 2. This pragmatic approach acknowledges that no single measurement perfectly captures GIST response biology.

Common Pitfalls to Avoid

• Don't rely solely on size measurements in the first 3-6 months of TKI therapy
• Don't stop effective therapy based on size increase alone without assessing density
• Don't ignore patient compliance and drug interactions as causes of apparent progression 2, 1
• Don't forget that response assessment requires experienced interpretation—early progression especially should be confirmed by expert review 1