Study objectives: To evaluate factors predicting the diagnostic yield of transbronchial biopsy (TBB) using endobronchial ultrasonography with a guide sheath (EBUS-GS) in small peripheral pulmonary lesions (PPLs) </= 30 mm in mean diameter.
Design: Retrospective analysis.
Patients and methods: One hundred fifty-five consecutive patients with 158 small PPLs underwent TBB using EBUS-GS.
Results: A definitive diagnosis was established by TBB using EBUS-GS in 106 PPLs (67%). The diagnostic yield of PPLs </= 15 mm in mean diameter (40%) was significantly lower than that of PPLs > 15 mm and </= 30 mm in mean diameter (76%; p < 0.001). PPLs in which the probe was positioned within the PPL on the endobronchial ultrasonography (EBUS) image had a higher diagnostic yield (83%) than PPLs in which the probe was positioned adjacent to the PPL (61%) or outside the PPL (4%; p < 0.001). There were no significant differences in diagnostic yield for underlying disease, location, CT scan bronchus sign, operator, or type of EBUS probe. In the multivariate analysis, only the position of the probe (within or adjacent to the PPL when judged against outside the PPL) was determined to be a significant factor predicting diagnostic yield. On the other hand, a pathologic diagnosis was established with the first, second, third, fourth, and fifth biopsy specimens in 65%, 80%, 87%, 91%, and 97% of PPLs, respectively.
Conclusions: The position of the probe (ie, within or adjacent to the PPL) is a significant factor in predicting the diagnostic yield of TBB using EBUS-GS for small PPLs; the optimum number of biopsy specimens is at least five.