What is the diagnostic yield of Endobronchial Ultrasound-guided Transbronchial Needle Aspiration (EBUS TBNA) with Rapid On-Site Evaluation (ROSE)?

Diagnostic Yield of EBUS-TBNA with ROSE

EBUS-TBNA with ROSE provides a diagnostic yield of approximately 78-92%, with ROSE offering potential advantages in reducing the number of needle passes and additional procedures, though not consistently improving the overall diagnostic accuracy compared to EBUS-TBNA without ROSE.

Overall Diagnostic Yield of EBUS-TBNA

• EBUS-TBNA provides safe and minimally invasive access to mediastinal and hilar lymph nodes with a pooled diagnostic accuracy of approximately 79% (95% CI, 71-86) across various conditions 1, 2
• Diagnostic yields range from 54% to 93% depending on the condition being evaluated and operator experience 1, 2
• For suspected sarcoidosis, EBUS-TBNA has a diagnostic yield of approximately 74.5-80%, which is significantly better than conventional TBNA (48.4%) 1, 3
• For tuberculosis, diagnostic yield ranges from 79.2% to 84.8% across studies 1

Impact of ROSE on Diagnostic Yield

• The 2025 CHEST guidelines suggest using ROSE for EBUS-TBNA in patients with suspected malignant disease (Conditional Recommendation, very low certainty of evidence) 1
• A pooled analysis of five observational studies showed diagnostic yield of 78.0% with ROSE versus 71.4% without ROSE (OR 2.35, CI 1.47-3.74) 1, 4
• Individual studies have reported diagnostic accuracies of:
  • 91% with ROSE versus 83% without ROSE (non-significant trend, p=0.08) 1
  • 100% with ROSE versus 86% without ROSE (p=0.005) 1
  • 92.1% with ROSE versus 89.2% without ROSE (p=0.272) 1
• However, a 2016 CHEST guideline meta-analysis found no statistical difference in diagnostic yield between studies using ROSE (80.1%) versus those without ROSE (81.3%) 1

Benefits of ROSE Beyond Diagnostic Yield

• ROSE significantly reduces the number of needle passes required:
  • 2.2 vs 3.1 passes (p<0.001) 1
  • 2.3 vs 4.0 passes (p<0.01) 1
  • 3.4 vs 6.1 passes (p<0.001) 1, 4
• ROSE decreases the need for additional diagnostic procedures in the same setting (11% vs 57%, p<0.001) 1
• ROSE reduces the percentage of non-diagnostic specimens (0.9% vs 4.4%, p=0.018) 1, 4
• ROSE helps with appropriate specimen triage for molecular testing and ancillary studies 5, 6

Optimal Sampling Technique

• In the absence of ROSE, a minimum of 3 separate needle passes per sampling site is recommended for optimal diagnostic yield 1
• Sample adequacy reaches 90.1% after the first pass, 98.1% after two passes, and 100% after three passes 1, 2
• Sensitivity for differentiating malignant from benign lymph nodes increases from 69.8% with one pass to 95.3% with three passes 1, 2

Factors Affecting Diagnostic Yield

• Operator experience significantly impacts diagnostic yield, with improvement noted after the first 15 procedures 3
• Cytopathologist experience also influences diagnostic accuracy, with improved interpretation over time 3, 6
• No significant difference in diagnostic yield has been observed between different needle gauges 1

Practical Considerations

• ROSE requires cytopathology personnel available near the procedure site, which may not be feasible at all centers 1, 5
• Alternative models for delivering ROSE include telecytopathology, trainees, cytotechnologists, and artificial intelligence 1
• The decision on collection medium should be discussed between the proceduralist and pathology department based on local practice and testing needs 1
• EBUS-TBNA with ROSE is particularly valuable when molecular testing is planned, as it helps ensure adequate tissue sampling 1, 6

Potential Limitations and Controversies

• Reimbursement models and pathology staffing limitations may restrict widespread implementation of ROSE 1
• Some studies challenge the notion that ROSE is beneficial for EBUS-TBNA specimens, showing no remarkable difference in diagnostic yield 7
• Cost-effectiveness must be considered, especially in resource-limited settings where conventional TBNA with ROSE may be a viable alternative 8