Objective: A frequent problem in patients with intrathoracic malignancies neighboring central airways is the question of whether the airway wall is infiltrated by the tumor or if it is merely compressed. This distinction can often not be made with certainty with the help of chest CT alone, but frequently necessitates surgical biopsy or exploration. We prospectively studied the utility of endobronchial ultrasound (EBUS) in this clinical circumstance.
Methods and patients: Between May 1999 and July 2000, 131 consecutive patients with central thoracic malignancies potentially involving the airways were enrolled into the study. Patients underwent chest CT followed by standard bronchoscopy together with EBUS and subsequent surgical evaluation. The bronchoscopists did not know the radiologist's interpretation of the chest CT before performing EBUS. The ability of chest CT and EBUS to distinguish between compression and infiltration was measured against the histologic results.
Results: One hundred five patients completed the trial by undergoing surgery. In 81 patients (77%), the CT scan was read as consistent with tumor invasion. EBUS only showed invasion in 49 cases (47%). Histology after surgery revealed a specificity of 100%, a sensitivity of 89%, and an accuracy of 94% for EBUS. Chest CT was far inferior, with a specificity of 28%, a sensitivity of 75%, and an accuracy of 51%.
Conclusion: We conclude that EBUS is a highly accurate diagnostic tool and superior to chest CT in evaluating the question of airway involvement by central intrathoracic tumors. In the hands of experienced endoscopists, EBUS may become the procedure of choice for this question.