Oncology Response Criteria: RECIST & PERCIST
Anatomic and metabolic response frameworks, and where each applies
Quick Answer
Tumor response is scored with the framework that matches the biology and modality. RECIST 1.1 is the anatomic (size-based) standard on CT/MRI. PERCIST 1.0 is the metabolic (FDG) standard, using SUL-peak of the hottest lesion against a liver reference. Disease- and therapy-specific frameworks refine these: Choi (GIST — density + size), iRECIST (immunotherapy pseudoprogression), Deauville/Lugano (lymphoma), and RECIP (PSMA-PET). Matching the right criterion — and comparing like-with-like — is what makes a response call valid.
Why Different Criteria
Size change (RECIST) can lag or misrepresent biology: GIST responds by becoming hypodense without shrinking, immunotherapy can cause transient pseudoprogression, and metabolic change often precedes anatomic change. Each framework encodes the response signal that best predicts benefit for its setting.
RECIST 1.1 (Anatomic)
- Target lesions: up to 5 measurable lesions (≤ 2 per organ); sum of longest diameters (short axis for nodes).
- Response categories:
| Category | Definition (summary) |
|---|---|
| Complete response (CR) | Disappearance of all target lesions |
| Partial response (PR) | ≥ 30% decrease in sum of diameters |
| Progressive disease (PD) | ≥ 20% increase (and ≥ 5 mm absolute) or new lesions |
| Stable disease (SD) | Neither PR nor PD |
Non-target lesions and new lesions are also assessed.
PERCIST 1.0 (Metabolic, FDG)
- Measures SUL-peak (SUV normalized to lean body mass) of the single hottest lesion, against a normal-liver reference (mean liver SUL + tolerance).
- Same scanner/protocol and uptake time across serial scans (metabolic change is highly technique-sensitive).
- Categories: complete metabolic response (CMR), partial metabolic response (PMR, ≥ 30% SUL-peak decrease), stable metabolic disease (SMD), progressive metabolic disease (PMD, ≥ 30% increase or new lesions).
Metabolic response frequently precedes anatomic (RECIST) change.
Disease/Therapy-Specific Frameworks
| Framework | Setting | Signal |
|---|---|---|
| Choi | GIST (imatinib) | ≥ 10% size or ≥ 15% CT-density decrease |
| iRECIST | Immunotherapy | Requires confirmation of progression (pseudoprogression) |
| Deauville / Lugano | Lymphoma | 5-point uptake vs blood pool/liver |
| RECIP 1.0 (PROMISE) | PSMA-PET (prostate) | Tumor-volume change + new lesions |
| PERCIST | Broad FDG oncology | SUL-peak metabolic response |
The Immunotherapy Caveat
Checkpoint-inhibitor therapy can produce pseudoprogression (transient enlargement/new inflammatory uptake before response) and immune-related FDG-avid inflammation. iRECIST addresses this by requiring confirmation of progression on a follow-up scan (unconfirmed progression = "iUPD"; confirmed = "iCPD"), preventing premature "failure" calls.
Common Pitfalls
- Mismatching the criterion to the setting (e.g. size-only RECIST for GIST, missing a Choi response).
- Comparing PERCIST across different scanners/protocols/uptake times (metabolic change is technique-sensitive — like PERCIST-forbidden).
- Calling immunotherapy progression without confirmation (pseudoprogression).
- Ignoring new lesions, which define progression across frameworks.
Board Pearls
Use the framework that fits the biology: RECIST 1.1 (anatomic size) on CT/MRI; PERCIST 1.0 (metabolic SUL-peak of the hottest lesion vs a liver reference) for FDG. Metabolic response often precedes anatomic change.
GIST needs Choi (density + size, not size alone); immunotherapy needs iRECIST (confirm progression — pseudoprogression); lymphoma uses Deauville/Lugano; PSMA-PET uses RECIP. Matching the criterion to the setting is half the job.
PERCIST demands like-with-like: same scanner/protocol/uptake time across serial scans (metabolic values are technique-sensitive), SUL normalized to lean body mass, and a normal-liver reference. New lesions define progression in every framework, and immune-related inflammation/pseudoprogression must be confirmed before calling failure.
Related Pages
- Reference: Lymphoma staging & response (Lugano/Deauville), PSMA-PET reporting (PROMISE/PSMA-RADS).
- Therapy: Theranostics response criteria (RECIP/PROMISE/PCWG3).
- Disease: Sarcoma & GIST (Choi), melanoma (immunotherapy response); tracer: FDG.
Figure / Diagram Suggestions
- A criterion selector by setting (RECIST / PERCIST / Choi / iRECIST / Deauville / RECIP).
- A metabolic-before-anatomic timeline (SUL falls before size).
- An iRECIST confirmation flow (iUPD → confirm → iCPD).
Self-Check (Board-Style)
Q1. A GIST on imatinib is unchanged in size but markedly hypodense. Which criterion captures the response, and how?
Answer: Choi criteria — a ≥ 10% size or ≥ 15% CT-density decrease counts as response; size-only RECIST would miss it.
Q2. What does PERCIST measure, and what reference organ anchors it?
Answer: The SUL-peak (lean-body-mass-normalized SUV) of the single hottest lesion, referenced to normal liver — with a ≥ 30% decrease defining partial metabolic response.
Q3. On immunotherapy, a lesion enlarges with new FDG-avid nodes at week 8. How should progression be handled?
Answer: Treat as unconfirmed (possible pseudoprogression/irAE) — iRECIST requires confirmation on a follow-up scan before calling confirmed progression.
Q4. Why must serial PERCIST studies use the same scanner and uptake time?
Answer: Metabolic (SUL) values are highly technique-sensitive (reconstruction, uptake time, calibration); comparing mismatched protocols invalidates the response call.
Evidence & sources
https://nuc.med/oncology-response-criteria-percist/ Report an issue