Comparative studies between radiological vs pathological Criteria for extra nodal extension
Reading File
Reading File
Finding Sources
Searching PubMed
"extranodal extension" AND radiology AND pathology
Searching PubMed
"extranodal extension" AND imaging
Searching the Web
radiologic pathological extranodal extension head neck cancer imaging criteria pathology correlation systematic review
Reading Article
PMID: 36071683
Reading Article
PMID: 33493691
Reading Article
PMID: 32949282
Reading Article
PMID: 29970315
Reading Article
PMID: 35315548
Reading Article
PMID: 38936388
Reading File
Assuming you mean head and neck squamous cell carcinoma (HNSCC), where ENE is best studied.
Key point: radiology and pathology are related but not interchangeable. Imaging detects macroscopic disruption/invasion around a metastatic node, whereas pathology confirms tumor beyond the nodal capsule, including microscopic ENE that imaging commonly misses.
| Domain | ENE criterion | Main limitation |
|---|---|---|
| Pathological ENE (pENE) | Viable metastatic tumor extends through the lymph-node capsule into surrounding soft tissue. Studies often subclassify minor ENE ≤2 mm and major ENE >2 mm beyond capsule. | Requires neck-dissection specimen, appropriate sampling, and consistent reporting. |
| Radiological ENE (rENE/iENE) | CT/MRI signs including irregular or indistinct nodal margins, capsular disruption, perinodal fat stranding, infiltration/loss of adjacent fat planes, matted nodes, or direct invasion of muscle, vessel, skin, or other structures. Central necrosis is supportive but is not specific. | Cannot reliably identify microscopic ENE and has variable thresholds between studies/readers. |
Cross-sectional imaging has poor sensitivity for microscopic ENE, although highly specific CT interpretations can be clinically useful (Cummings Otolaryngology, p. 1894).
Comparative radiology-pathology studies
| Study | Population / reference | Main comparative result |
|---|---|---|
| Almulla et al., 2018 (PMID 29970315) | 508 oral cavity SCC patients; preoperative CT/MRI compared with pENE after neck dissection | For imaging-to-surgery interval ≤8 weeks: rENE sensitivity 52%, specificity 96%, PPV 93%, NPV 66%. CT was more accurate than MRI (80% vs 63%). Accuracy declined to 48% if surgery was >8 weeks after imaging. |
| Blasco et al., 2022 (PMID 35315548) | 334 pN+ oral cavity SCC patients; rENE versus minor/major pENE | Overall rENE sensitivity 37%, specificity 98%. Sensitivity improved only to 48% for major pENE >2 mm, showing that even larger histologic ENE may be occult on imaging. |
| Morey et al., 2022 systematic review (PMID 36071683) | HPV-positive oropharyngeal SCC | CT pooled sensitivity 77%, specificity 60%. PET/CT: sensitivity 37.5%, specificity 97%. Combined CT/MRI: sensitivity 62%, specificity 78%. Evidence was limited to six eligible studies. |
| Park et al., 2021 diagnostic meta-analysis (PMID 32949282) | 22 studies, 2,478 HNSCC patients | CT: sensitivity 73%, specificity 83%. MRI: sensitivity 60%, specificity 96%. Necrosis increased sensitivity (81% vs 51%), while adjacent-plane infiltration increased specificity (94% vs 65%). CT specificity was lower in HPV-positive OPSCC. |
| Abdel-Halim et al., 2021 systematic review/meta-analysis (PMID 33493691) | 25 studies, 3,391 patients; histopathology reference standard | CT sensitivity/specificity 76%/77%; MRI 72%/78%; PET/CT 80%/83%. PET/CT had higher sensitivity than CT and MRI, but overall diagnostic differences were not statistically significant. |
Practical comparison of the criteria
Imaging features that most closely predict pathological ENE
-
Definite invasion of adjacent structures or loss of tissue planes
Most specific imaging sign. CT/MRI can identify gross spread into muscle, vessel, skin, or other neighboring structures. -
Irregular nodal margin with perinodal fat infiltration/stranding
Common rENE criterion, but less reliable for small-volume pENE and may be confounded by inflammation or post-biopsy change. -
Matted/coalescent nodes
Suggests advanced regional disease and may indicate ENE, but is not a direct histological equivalent. -
Central nodal necrosis
Raises suspicion and tends to improve sensitivity, but should not alone be labeled ENE, particularly in HPV-associated cystic nodal metastases.
Why radiology-pathology discordance occurs
- Microscopic pENE can extend only a small distance beyond the capsule and is below CT/MRI spatial resolution.
- Imaging is often evaluated node-by-node, while pathology may be reported by nodal specimen or neck level, complicating direct matching.
- Delay between imaging and surgery allows disease progression, with one large oral-cavity series showing poorer agreement beyond 8 weeks.
- Histopathologic definitions, tissue processing, sampling density, and use of minor/major ENE cutoffs vary.
- HPV-positive OPSCC often has cystic/necrotic nodal metastases that can mimic imaging ENE.
Current reporting approach
The 2024 Head and Neck Cancer International Group consensus recommends harmonized terminology and standardized imaging criteria for iENE, but also states that its proposed grading system still requires validation (Henson et al., 2024). In a report, separate the observation from certainty:
- No iENE: smooth nodal margin, preserved surrounding fat planes.
- Suspicious iENE: irregular margin and limited adjacent fat-plane stranding.
- Definite/advanced iENE: unequivocal extension into neighboring soft tissue or structures.
Avoid wording such as “pathological ENE present” on imaging. Prefer: “Imaging findings suspicious for/consistent with extranodal extension; pathological confirmation is required when surgical tissue is available.”
Bottom line
- pENE remains the reference standard.
- rENE has high specificity when there is gross invasive disease, so a positive study can be clinically meaningful.
- rENE has limited sensitivity, particularly for microscopic and minor ENE. Therefore, a negative CT, MRI, or PET/CT does not exclude pENE.
- CT is usually the pragmatic first-line modality; MRI may offer high specificity; PET/CT may be somewhat more sensitive in pooled analyses, but no modality reliably replaces pathology.
Recent evidence has not overturned this conclusion, but it emphasizes that both imaging and histopathology need standardized terminology and criteria.
This is a shared conversation. Sign in to Orris to start your own chat.