Radial endobronchial ultrasound-guided transbronchial biopsy for peripheral pulmonary malignancy: biopsy- or brushing-first?

Diagnosis of peripheral pulmonary malignancy can be attained via a variety of modalities, such as bronchoscopy, computed tomography (CT)-guided transthoracic needle biopsy and surgery. In the past two decades, advances in bronchoscopic procedures, such as radial endobronchial ultrasound (R-EBUS), virtual bronchoscopic navigation and electromagnetic navigation have made transbronchial biopsy (TBB) a more appealing and favorable approach [1, 2]. Of the above-mentioned techniques, R-EBUS-guided TBB provides a fair diagnostic yield and an excellent safety signal in diagnosing peripheral pulmonary malignancy [3‐5]. In addition to TBB, auxiliary diagnostic tools, such as bronchial brushing and washing, are commonly used to reach the cytologic diagnosis, and combining these procedures may achieve a higher diagnostic yield than TBB alone [6, 7].

Besides the diagnostic modalities and tools, the details of the bronchoscopic procedures may affect their diagnostic yield. For instance, in bronchoscopically visible lesions, a brushing-first strategy provided a significantly higher diagnostic yield for lung cancer than a biopsy-first strategy [8]. In a similar clinical context, bronchial washing performed before or after endobronchial biopsy did not affect the diagnostic yield of biopsy and washing [9]. Regarding R-EBUS-guided TBB of peripheral pulmonary malignancy, both biopsy- and brushing-first strategies may be applied in clinical practice [10, 11]. However, to the best of our knowledge, no studies have been conducted to assess the ideal sequence of TBB and bronchial brushing during R-EBUS-guided procedures.

Therefore, the aims of this study were to investigate whether a biopsy-first or brushing-first strategy confers a better diagnostic yield and safety signal for R-EBUS-guided TBB of peripheral pulmonary malignancy.

The present work is the first study to compare the diagnostic yield and complications of R-EBUS-guided TBB plus brushing for peripheral pulmonary malignancy using a biopsy-first or brushing-first strategy. The main findings of our study are as follows: (a) the overall diagnostic yield of TBB plus brushing of peripheral pulmonary malignancy was 73%; (b) a biopsy-first strategy provided diagnostic sensitivity for malignancy similar to a brushing-first strategy; (c) solid lesion appearance and position of the R-EBUS probe within the lesion were two features significantly associated with the diagnostic yield of TBB plus brushing; (d) the overall complication rate was 6.8%, with hemorrhage being the most common complication; (e) the safety signal did not differ between the biopsy-first and brushing-first groups.

The most important finding in this study is that the two strategies resulted in a comparable diagnostic yield for R-EBUS-guided procedures for peripheral pulmonary malignancy. In fact, most of the previous reports regarding TBB and bronchial brushing with the assistance of R-EBUS did not specify the sequence of the procedures [6, 16‐18]. Kurimoto et al. adopted a brushing-first strategy for TBB and reported a high detection rate of 81% for peripheral pulmonary malignancy [11]; on the other hand, Roth et al. performed TBB prior to bronchial brushing and the diagnostic yield for malignancy was less than 50% [7]. A recent study by Hou et al. examined the optimal sequence of forceps biopsy and bronchial brushing of visible endobronchial lung cancer and found a significantly higher diagnostic rate (87%) in the brushing-first group compared to the biopsy-first group (79%) [8]. The authors rationalized their results by suggesting that endobronchial tumors can be more difficult to identify and sample after biopsy, since hemorrhage may contaminate the field available for brushing [8]. In contrast to previous experience with bronchoscopically visible lesions, our study showed a similar diagnostic yield for peripheral pulmonary malignancy in both the biopsy-first and brushing-first groups. One speculation is that with the guidance of R-EBUS, bronchial brushing of peripheral pulmonary lesions can be performed without visual aids, and post-biopsy bleeding will not obscure their visualization; thus, the procedure may be performed either before or after TBB. The other is that in experienced hands [4], a high quality R-EBUS-guided procedure could be performed to achieve a superior diagnostic sensitivity to peripheral pulmonary malignancy in both biopsy-first and brushing-first groups.

Our diagnostic sensitivity for peripheral pulmonary malignancy using TBB was consistent with that of previous studies of R-EBUS for malignant lesions, in which the diagnostic yield of R-EBUS ranged from 47 to 81% [6, 11, 16, 19‐22]. The prevalence of malignancy present in the population being studied is a well-recognized explanation for the observed differences in the diagnostic sensitivity of R-EBUS-guided TBB for peripheral pulmonary lesions [5]. However, a wide variation in the yield of TBB still exists, even though only a subset of malignant lesions are chosen for evaluation in the literature. This heterogeneity may be explained by the discrepancies in other characteristics related to the cases, such as lesion size, personnel and institutional experience, and concomitant use of additional tools, like ROSE. The lowest diagnostic yield of 47% was observed in a study that included only solitary pulmonary nodules smaller than 20 mm [22], and undoubtedly, procedural experience improves the performance of R-EBUS-guided TBB [4]. Moreover, the application of ROSE has been shown to improve the diagnostic yield of TBB under R-EBUS guidance [23]. Therefore, it is important to consider patient and staff factors, and auxiliary modalities used when assessing how a certain technology like R-EBUS performs in clinical practice.

In a recent study by Chen, et al. [24], the diagnostic yield of R-EBUS-guided procedures for peripheral pulmonary lesions was positively correlated with the lesion size. Also, a meta-analysis found that lesion size was a significant determining feature in TBB performance [25]. However, the present study did not show such a finding. Our study population was composed solely of those with malignant pathology, and the procedures were conducted by a well-developed and experienced team. These specific settings may partly explain the discrepancy between our findings and others.

Safety is certainly a concern when choosing a modality or procedure. Consistent with prior studies [3, 26], a favorable safety profile with no mortality or sequelae with the use of R-EBUS-guided TBB for peripheral pulmonary malignancy was observed in this work. Pneumothorax is a well-known and potentially catastrophic complication after TBB, and our occurrence rate of 1.4% lies on the low end of those previously reported: 0 to 5.1% [3, 6, 10, 11, 17, 20‐22, 26]. Bleeding is another major complication of TBB; however, its incidence is more difficult to compare across studies given the wide variation in definitions. On average, an incidence rate of 0.7% was reported for procedure-related bleeding regardless of its definition [11, 16, 17, 20‐22, 26‐28], but no major or serious bleeding events occurred with any patient in these reports. The bleeding rate observed in this study, 4.8%, seems to be higher than those of prior studies, but we adopted a stricter criterion for reporting this complication (any bleeding requiring further intervention) than have other studies (bleeding necessitating premature termination of the procedure or a bleeding amount of ≥30 ml) [11, 26]. With regard to our study aim of comparing safety signals between the biopsy-first and brushing-first strategies, the complication profiles and rates for R-EBUS-guided procedures were similar between the two groups of patients. Thus, the two procedure strategies studied herein were indistinguishable in terms of both efficacy and adverse events for patients with peripheral pulmonary malignancy.

In line with previous studies [4, 10, 11, 17, 18], ours found that probe position is a major determinant of the diagnostic yield of R-EBUS-guided TBB for peripheral pulmonary malignancy. This finding reinforces the importance of navigating the probe to the desired position that is visualized within the target lesion. In this study, the R-EBUS probe was more likely to be placed within the malignancy in patients in the brushing-first group, so the difference between probe positions may complicate the comparison of the diagnostic efficacy of the two study groups before adjusting for this confounder. However, in both the multivariate model and the subgroup analysis after taking probe position into consideration, the diagnostic sensitivity for peripheral pulmonary malignancy remained similar between the two strategies. This suggests the robustness of our study findings.

A couple of caveats pertaining to this study should be mentioned. First, the study findings represent a single-center experience with R-EBUS-guided TBB for peripheral pulmonary malignancy carried out by a well-developed bronchoscopy team; thus, the results may not be generalized to other settings, such as a less-experienced institution or a lower-level healthcare facility. Nonetheless, as the first study to deal with this issue, we hope our report will encourage more large-scale and elaborate studies in this field. Second, our study was retrospective, and as such, the choice of procedure strategy was at the discretion of the in-charge pulmonary faculty. In this regard, a selection bias may exist; however, baseline features of the participants were nearly balanced between the biopsy-first and brushing-first groups, and further statistical adjustment did not identify any significant differences in the study outcomes between the two groups.

In summary, the timing of bronchial brushing, either before or after biopsy, for peripheral pulmonary malignancy, at least in experienced hands, did not influence the diagnostic yield of R-EBUS-guided TBB plus brushing. Moreover, a similarly favorable safety signal was observed between the two strategies. Therefore, current evidence indicates that in patients with peripheral pulmonary lesions suspected of being malignant, either biopsy-first or brushing-first would be a viable and acceptable strategy with respect to diagnostic sensitivity and safety during R-EBUS-guided procedures.