周邊型肺部病灶：支氣管鏡檢中結合錐狀束電腦斷層衍生之強化螢光透視與支氣管刷檢的全面評估

Abstract

隨著低劑量電腦斷層（low-dose computed tomography, LDCT）於肺癌篩檢的逐漸普及，周邊型肺部病灶（peripheral pulmonary lesions, PPLs）的發現數量也隨之增加，亟需有效的切片方式。目前常見的切片方法包括經電腦斷層導引的經皮切片，以及運用氣管鏡的經支氣管切片（transbronchial biopsy, TBB）。支氣管鏡超音波（radial endobronchial ultrasound, rEBUS）能提升病灶定位能力，但取樣過程中缺乏即時影像。錐狀束電腦斷層衍生的強化螢光透視檢查（cone-beam computed tomography-derived augmented fluoroscopy, CBCT-AF）則整合精確影像與即時螢光導引，可望提高rEBUS引導切片的成效，尤其是小病灶，但目前相關的研究仍然有限。此外，以傳統氣管鏡取得的檢體量常不足以進行非小細胞肺癌（non-small cell lung cancer, NSCLC）的分子檢測。透過rEBUS進行的支氣管刷檢或許能提供額外的檢體，用於細胞病理診斷及表皮生長因子受體（epidermal growth factor receptor, EGFR）基因檢測，然而其實際效果仍需進一步研究。 本研究分為兩部分。第一部分旨在確定CBCT-AF與rEBUS的結合使用是否比單獨使用rEBUS能提高PPLs的診斷率。患有PPLs的患者將接受CBCT-AF結合rEBUS或僅單獨使用rEBUS的比較分析。我們假設CBCT-AF將顯著改善病灶定位，從而提高診斷準確性，同時具有與單獨使用rEBUS相似的安全性。第二部分旨在評估支氣管刷檢在細胞病理學和分子檢測中的可行性，特別是在NSCLC患者中檢測EGFR突變的效用。我們假設支氣管刷檢適合進行EGFR突變檢測，從而減少因樣本不足而需進行額外侵入性檢查的需求。 第一部分研究納入自2018年10月至2019年7月間，接受rEBUS導引經支氣管切片的周邊型肺部病灶患者，探討在EBUS檢查中結合CBCT-AF之成效與安全性。透過傾向分數1:1配對後，最終分析53對患者。研究結果顯示，rEBUS結合CBCT-AF組的診斷率明顯優於僅使用EBUS組（75.5% vs. 52.8%，p=0.015），尤其在病灶小於或等於3公分時，更能顯著提升診斷率（73.5% vs. 36.1%，p=0.002）。兩組並無顯著併發症差異（3.8% vs. 5.7%，p=1.000）。此外，在rEBUS結合CBCT-AF組中，約45.3%的病灶在傳統螢光透視下無法可視化，但透過CBCT-AF仍能定位並順利進行標註；整個操作流程、CBCT 與螢光透視的平均輻射劑量分別為 19.59、16.4 和 3.17 Gy cm²。總結而言，第一部分研究顯示，將CBCT-AF與rEBUS 結合用於周邊肺病灶之經支氣管切片，能提供相對令人滿意的診斷率，且所需操作時間比未使用CBCT-AF的組別更短。此外，該方法在操作相關的併發症及輻射曝露劑量方面亦被證實是安全的。第二部分研究聚焦於周邊肺癌患者接受支氣管鏡超音波檢查時，採用支氣管刷檢檢體在細胞病理及分子檢測，特別是EGFR基因突變定序上的應用價值。研究共收錄941位周邊型肺部病灶患者，最終有624人確診為非鱗狀細胞肺癌，其中376人（60.3%）在氣管鏡刷檢檢體中得到陽性細胞學結果。顯示若胸部電腦斷層可見支氣管徵象(bronchus sign)、或rEBUS探頭能位於病灶內部，則刷檢診斷率更高。在這些陽性檢體中，96.5%可成功進行EGFR基因定序，且與配對組織樣本的突變檢測一致性達88.7%（kappa=0.745，p<0.001）。值得注意的是，有144位患者最初透過經支氣管切片無法完成病理診斷或EGFR檢測，但其中57位（39.6%）可藉由同次刷檢獲得所需檢體。整體而言，rEBUS導引支氣管刷檢可作為輔助性取樣方式，提供周邊肺癌診斷及EGFR突變檢測的額外可靠途徑。 綜觀本研究兩部分結果,對於接受支氣管鏡檢查之周邊型肺部病灶患者,應用CBCT-AF輔助rEBUS不僅能提升診斷率與縮短操作時間,亦可維持良好安全性。當患者最終確診為非鱗狀非小細胞肺癌時,rEBUS之支氣管刷檢亦能提供額外且充足的檢體以進行EGFR突變檢測,減少重複侵入性檢查之需求。本研究結果可作為未來臨床診斷與治療決策的重要參考依據。

With increased low-dose computed tomography screening for lung cancer, more peripheral pulmonary lesions (PPLs) are detected, emphasizing the need for effective biopsy methods. Traditional approaches include CT-guided percutaneous biopsy and bronchoscopic transbronchial biopsy (TBB). Radial endobronchial ultrasound (rEBUS) improves lesion localization but lacks real-time imaging during sampling. Cone-beam computed tomography-derived augmented fluoroscopy (CBCT-AF) combines precise imaging and real-time fluoroscopy, potentially enhancing rEBUS biopsy, particularly for small lesions, though comparative data are limited. Additionally, limited biopsy samples often hinder molecular analysis in non-small cell lung cancer (NSCLC). Bronchial brushing during rEBUS may provide sufficient material for diagnosis and genetic testing, such as epidermal growth factor receptor (EGFR) mutation detection, but this approach requires further evaluation. This study has two parts. First, we evaluated whether combining CBCT-AF with rEBUS increases the diagnostic yield for PPLs compared to rEBUS alone, hypothesizing improved lesion localization and diagnostic accuracy without compromising safety. Second, we assessed the feasibility of bronchial brushing specimens for cytological diagnosis and EGFR mutation analysis in NSCLC, hypothesizing this approach could reduce the need for additional invasive procedures. The first part of this study enrolled patients with PPLs undergoing rEBUS-guided TBB from October 2018 to July 2019, evaluating the efficacy and safety of combining CBCT-AF with rEBUS. After 1:1 propensity score matching, 53 matched pairs were analyzed. Results demonstrated a significantly higher diagnostic yield in the rEBUS combined with CBCT-AF group compared to rEBUS alone (75.5% vs. 52.8%, p=0.015), particularly in lesions ≤3 cm (73.5% vs. 36.1%, p=0.002). Complication rates were similar between groups (3.8% vs. 5.7%, p=1.000). Additionally, approximately 45.3% of lesions in the CBCT-AF group were not visible on conventional fluoroscopy but could still be successfully localized and labeled through CBCT. The mean radiation doses for the entire procedure, CBCT, and fluoroscopy were 19.59, 16.4, and 3.17 Gy·cm², respectively. Overall, combining CBCT-AF with rEBUS provided an improved diagnostic yield, reduced procedure time, and was proven safe regarding complications and radiation exposure. The second part assessed the diagnostic and molecular utility of bronchial brush samples obtained during rEBUS in patients with peripheral lung cancer, specifically for EGFR mutation analysis. Of 941 PPL patients enrolled, 624 were diagnosed with non-squamous NSCLC. Positive cytology from bronchial brush samples was obtained in 376 patients (60.3%), particularly when a bronchus sign was visible on computed tomography or the rEBUS probe was located within the lesion. Among these positive samples, 96.5% successfully underwent EGFR mutation sequencing, showing an 88.7% concordance rate with paired tissue samples (kappa=0.745, p<0.001). Notably, among 144 patients initially negative by TBB for pathology or EGFR analysis, 57 (39.6%) obtained adequate samples through bronchial brushing. Thus, rEBUS bronchial brushing serves as a reliable supplementary sampling method for diagnosing peripheral lung cancer and determining EGFR mutation status. In conclusion, this two-part study shows that in patients with PPLs undergoing bronchoscopy, the application of CBCT-AF in conjunction with rEBUS can enhance diagnostic yield and shorten procedural time while maintaining a high degree of safety. Moreover, for patients ultimately diagnosed with non-squamous NSCLC, rEBUS bronchial brushing offers an additional, sufficient sample source for EGFR mutation testing, thereby reducing the need for repeat invasive procedures. These findings can serve as an important clinical reference for future diagnostic strategies and treatment planning.