Accuracy of Inspiratory and Expiratory Sagittal CT Images for Diagnosing Tracheomalacia
Inspiratory and expiratory CT imaging is highly accurate for diagnosing tracheomalacia, with forced expiratory CT being the gold standard imaging modality that demonstrates superior sensitivity compared to end-expiratory imaging, though sagittal reformations specifically add limited diagnostic value beyond standard axial images.
Primary Imaging Protocol
The 2024 ACR Appropriateness Criteria establishes chest CT without IV contrast with paired inspiratory and forced expiratory imaging as the primary noninvasive imaging approach for tracheomalacia 1. This protocol has been widely adopted because:
- Forced expiratory CT is significantly more sensitive than end-expiratory CT, eliciting mean tracheal collapse of 62% versus only 17% at end-expiration 1
- In patients with suspected expiratory central airway collapse (ECAC), forced expiratory imaging showed >50% collapse at 38% of assessed levels compared to only 13% at end-expiratory CT 1
- Dynamic expiratory CT performs equally well with low-dose technique (40-80 mAs) compared to standard dose (240-280 mAs), maintaining high confidence in tracheal lumen measurement 2
Diagnostic Thresholds and Accuracy
The diagnostic criteria have evolved based on research findings:
- Current threshold: ≥70% airway collapse is now recommended, as >50% collapse was found in 78% (40/51) of healthy volunteers, making the older 50% threshold too sensitive 1
- The technique can change treatment approach in 41% of patients (12/29) who had previously undergone bronchoscopy alone, including modifications in surgical planning 1
Role of Sagittal Reformations
While volumetric CT allows sagittal reformations, their specific diagnostic contribution is modest:
- Sagittal reformations provided additional diagnostically significant findings (grade 3) in only 9% (2/22) of cases, with most additional findings being of no diagnostic significance 3
- The primary value of sagittal images relates to assessing relationships between masses/nodules and anatomic structures, not specifically for tracheomalacia diagnosis 3
- Axial images remain the primary plane for measuring tracheal collapse, as the standard measurement involves calculating percentage change in cross-sectional area between inspiratory and expiratory phases 4, 2
Specific Morphologic Signs
On inspiratory imaging, tracheal morphology is typically normal (94% of cases), with the "lunate" or "lemon" shape (elongated transverse diameter) being highly specific but rare, present in only 6% of tracheomalacia patients 5. On expiratory imaging, the "frown sign" (crescenteric collapse) is observed in 53% of patients with tracheomalacia and has practical value for detection when patients inadvertently breathe during routine CT 5.
Comparison to Bronchoscopy
Recent evidence demonstrates:
- 4D dynamic CT shows 91.6% diagnostic accuracy for tracheomalacia with 90.9% sensitivity and 100% specificity, with almost perfect agreement (κ = 0.913) compared to bronchoscopy 6
- However, inter-rater reliability is challenging for both modalities, with substantial agreement for bronchoscopy (AC2 = 0.79) and 4D-CT (AC2 = 0.66), indicating that distinguishing tracheomalacia subtypes remains difficult even for experts 7, 8
Clinical Pitfalls
Critical caveat: Only 1.5% of clinicians correctly diagnosed all cases when reviewing CT images, and 18.6% identified fewer than half correctly, with poor inter-rater agreement (κ = 0.21) 8. This highlights that while the imaging technique is accurate, interpretation requires significant expertise. Predictors of accuracy include ≥10 years in practice and pulmonology specialty training 8.
The presence of stridor during CT acquisition is positively associated with detecting abnormalities, suggesting that symptomatic imaging may improve diagnostic yield 6.