ILD-RADS Classification System
ILD-RADS (Interstitial Lung Disease Reporting and Data System) is a standardized CT reporting framework that categorizes interstitial lung disease into four categories (ILD-RADS 1-4) based on the likelihood of usual interstitial pneumonia (UIP) pattern, designed to improve diagnostic consistency and communication between radiologists and clinicians. 1, 2
The Four ILD-RADS Categories
The classification system stratifies findings into the following categories based on HRCT patterns:
ILD-RADS 1 (Typical UIP): CT features are diagnostic of UIP pattern, showing the classic combination of subpleural and basal-predominant reticular abnormalities, honeycombing, and traction bronchiectasis without features suggesting an alternative diagnosis 1, 3
ILD-RADS 2 (Probable/Possible UIP): CT features are suggestive but not definitive for UIP, typically showing reticular abnormalities and traction bronchiectasis but lacking definite honeycombing 1, 3
ILD-RADS 3 (Indeterminate for UIP): CT features are nonspecific and could represent either UIP or non-UIP patterns, requiring additional clinical correlation or tissue diagnosis 1, 3
ILD-RADS 4 (Inconsistent with UIP/Non-UIP): CT features are most consistent with a non-UIP diagnosis, such as NSIP, organizing pneumonia, or other specific ILD patterns 1, 3
Structured Reporting Elements
The ILD-RADS template systematically evaluates both pulmonary and extrapulmonary findings:
Pulmonary findings assessed include: lung volume, reticulations, traction bronchiectasis, honeycombing, nodules, cysts, ground-glass opacities, consolidation, mosaic attenuation, and emphysema 1, 3
Distribution patterns are characterized by:
- Axial distribution: central, peripheral, or diffuse 1
- Zonal distribution: upper, middle, or lower lung zones 1
Extrapulmonary findings include mediastinal abnormalities, pleural disease, tracheal changes, and bone/soft tissue lesions 1, 3
Complications specifically evaluated are acute infection, acute exacerbation, and malignancy 1
Clinical Utility and Performance
ILD-RADS demonstrates excellent clinical utility with significant association between assigned categories and final clinical diagnoses, particularly for ILD-RADS 1 and 4 categories 4. Studies show 50-57% of IPF patients are correctly assigned ILD-RADS 1, while 46-49% of non-IPF patients receive ILD-RADS 4 4.
The system shows excellent intra-reader agreement (κ = 0.873) and moderate inter-reader agreement (κ = 0.440) overall, with the highest inter-reader agreement for ILD-RADS 1 (κ = 0.681) and ILD-RADS 4 (κ = 0.481) categories 4, 3.
Integration with Multidisciplinary Diagnosis
ILD-RADS serves as a communication tool but does not replace the requirement for multidisciplinary discussion (MDD) integrating pulmonology, radiology, and pathology expertise 5, 6. The standardized lexicon facilitates more precise communication during MDD, which remains the gold standard for ILD diagnosis 5, 6.
Critical Distinction from ATS/ERS Classification
ILD-RADS is a radiologic reporting system, not a comprehensive disease classification—it focuses specifically on CT pattern recognition for UIP versus non-UIP differentiation 1, 2. The broader ATS/ERS classification system encompasses seven major idiopathic interstitial pneumonias (IPF, NSIP, RB-ILD, DIP, COP, AIP, LIP) plus rare IIPs and behavioral classifications 5, 6.
Practical Implementation
Radiologists report positive acceptance of ILD-RADS for daily practice, citing improved standardization and clarity in reporting 4. The template-based approach ensures comprehensive evaluation of all relevant imaging features while providing a clear categorical assessment that guides clinical decision-making 2, 4.
The key pitfall to avoid is using ILD-RADS categories as definitive diagnoses without clinical correlation—these categories indicate radiologic patterns that must be integrated with clinical, serologic, and when necessary, histopathologic data through formal MDD 5, 6, 4.