Applicability of Endobronchial Ultrasound and Virtual Bronchoscopic Navigation Combined with Rapid On-Site Evaluation in Diagnosing Peripheral Lung Lesions

Background To investigate the value of endobronchial ultrasound (EBUS) and virtual bronchoscopic navigation (VBN) combined with rapid on-site evaluation (ROSE) in diagnosing peripheral pulmonary lesions (PPLs). Methods Between January 1st 2019 to September 1st 2021, EBUS and VBN examination were performed in 250 consecutive patients with PPLs who were admitted to Zhangzhou Aliated Hospital of Fujian Medical University (Fujian, China). Finally, 198 eligible patients were randomly divided into ROSE group (100 cases) and non-ROSE group (98 cases). The diagnostic yield of brushing and biopsy, the complications, the procedure time, the diagnosis time and expense during diagnosis were analyzed.


Conclusion
In combination with ROSE, EBUS could signi cantly improve the positive rate of diagnosing PPLs, shorten diagnosis time and reduce expense during diagnosis. ROSE will be of great importance in the diagnosis of PPLs and medical resource.

Background
Peripheral pulmonary lesions (PPLs) refer to lesions that are located in the subsegmental bronchi and cannot be visualized directly by bronchoscopy [1]. With the widespread application of computer tomography (CT), PPLs have been increasingly detected [2]. It is becoming increasingly apparent that transbronchial biopsy (TBB) has become an important method for obtaining specimen from PPLs, however, the diagnostic yield widely ranges from36 to 76% [2][3][4][5]. The diagnosis of PPLs is still di cult because of its anatomic location far from segmental bronchus, which is unable to reach the lesion by routine bronchoscopy [3]. Particularly, pathological diagnosis was clinically important for those patients with benign PPLs including tuberculosis and pulmonary fungal diseases that can be distinguished from malignant lesions, which could avoid unnecessary operations and reduce medical expenses. Indeed, lung cancer as the most common PPLs is the leading cause of cancer-related death worldwide [4]. In recent years, radial endobronchial ultrasound (R-EBUS) emerged as a powerful tool during TBB and brushing in PPLs, and it led to the improvement of the diagnostic yield [5].
On the other hand, rapid on-site evaluation (ROSE) introduced by Park could help to quickly evaluate the satis ed specimen, form a preliminary diagnosis and guide the TBB operation in real time [6]. An emerging body of evidence indicates that the use of ROSE during endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) is a minimally invasive and highly accurate modality for the diagnosis of lymph node metastasis in lung cancer [7,8]. Based on ROSE, EBUS-TBNA signi cantly reduced the number of needle passes and complication rates [9], which can contribute to cost savings in the medical system [10]. Actually, a high concordance rate was reported between ROSE and histologic diagnosis [11].
However, little was known about whether the utility of ROSE can affect the diagnostic yield of EBUS TBB in PPLs [12][13][14]. Subject to the relatively small size of sample [12] and the cohort design [13,14], so the inference should be interpreted with coution. Additionally, the current data regarding whether several factors such as location and size of lesion [15,16], inadequate specimen collection affecting the diagnostic yields of EBUS TBB in PPLs [17] remain con rmed.
The aim of this study was to evaluate the value of R-EBUS in combination with ROSE in diagnosing PPLs and explore factors that can in uence the diagnostic yields.

Participants
This was a prospective randomized, controlled trial (RCT The study was approved by the ethical committee of Zhangzhou A liated Hospital of Fujian Medical University (ethics approval no. Zzsyy-2017-1116), and all patients provided informed written consent.

Biopsy procedure by EBUS and VBN
The location of the bronchus leading to the lesion was designed by VBN in advance (Ziostation2;Ziosoft Ltd, Tokyo, Japan; LungPoint; Bronchus Ltd, Mountain View, CA, USA; or DirectPath, Olympus Ltd, Tokyo, Japan). Bronchoscopy was performed applying a beroptic bronchoscope in combination with the R-EBUS (20MHz mechanical-radial type, UM-S20-20R or UM-S20-17S; Olympus, Japan) and guide sheath (GS) kit (K-201 or K-203; Olympus, Japan). The scope was inserted through the oral route, and each procedure was performed under local anesthesia with intravenous administration of midazolam for mild sedation. X-ray uoroscopy (VersiFlex VISTA, Hitachi, Japan) was applied to guide the insertion of the R-EBUS probe with GS through the working channel of the bronchoscope until the target site was reached.
After determining the location of the R-EBUS probe and GS within a target lesion, brushing and TBB cytology were perfromed for specimen collection. When the R-EBUS probe was adjacent to or outside the target lesion, the bronchus closest to the PPLs was meticulously searched under uoroscopy prior to collecting specimen. X-ray uoroscopy guidance was applied during biopsy and brushing sampling, as well as during removal of the GS after sampling. The procedure time was measured based on the interval between insertion and removal of the bronchoscope through the vocal cords. The diagnosis time was measured based on the interval between the R-EBUS operation to nal pathological diagnosis. And the expense during diagnosis indicated the costs between the R-EBUS operation to nal pathological diagnosis.

Rapid on-site specimen evaluation
The material obtained from bronchoscopic biopsy or brushing, was immediately expressed onto numbered glass slides. Diff stain was applied for specimen staining (Diff-Quik; Sysmex Ltd. Kobe, Japan). The stained slide was screened by the same experienced cytopathologist, who continuously reported the ndings in real time and announced when su cient diagnostic material had been obtained for a provisional diagnosis. The bronchoscopist modi ed or terminated the sampling process according to the information provided by the cytopathologist. Then tissue biopsy samples were placed in 10% formalin and were embedded in para n for routine histologic evaluation on hematoxylin and eosin staining.
Positive diagnostic criteria for ROSE: (1) The ROSE cytology showed cancer cells and nuclear heterogeneous cells; (2) The ROSE reported neutrophils, and lesions absorpted after anti-infective treatment; (3) The ROSE showed lymphocytes and other in ammatory cells.
For those non-diagnostic patients, the nal diagnosis was determined through additional medical examinations such as CT-guided percutaneous lung biopsy, surgical biopsy or anti-infection, antituberculosis therapy and follow-up for at least 6 months. No signi cant difference in lesion size, location of lesion and composition of disease was observed between both groups. 2. The impact of ROSE on the diagnostic yield of R-EBUS brushing and biopsy Table 2 indicates that the positive rate of brushing in the ROSE group is 68%, and the control group is 44%. The positive rate of biopsy in the ROSE group is 84%, and the controls is 74%. The differences in the diagnostic yield of R-EBUS brushing and biopsy between both groups were signi cant. 3. The Impact Of Rose On Bronchoscopy Complications Table 3 shows 2 cases of severe bleeding in ROSE group, and there is no signi cant difference in the incidence of bronchoscopy complications between both groups.     Normally distributed data were expressed as mean±SD, using a Student's t test for comparison. Differences were considered to indicate signi cance if a p value was <0.05. ROSE: rapid on-site evaluation.
6. The impact of ROSE on diagnostic yield related to location and size of the lesion between both groups Table 6 shows that the diagnostic yield of TBB in right upper lobe in the ROSE group is signi cantly higher than the controls. The signi cant difference was also observed in the size of lesion ≤2 cm in diameter.

Discussion
Notably, the 5-year survival rate after diagnosis of lung cancer was only 18% in the United States, so early diagnosis of malignant PPLs was also of great signi cance [19]. Meanwhile, it is currently believed that it may be clinically important for those patients with benign PPLs distinguished from malignant lesions.
Identi ed as an useful approach of PPLs, it was reported that the positive rate of percutaneous lung biopsy could reach 74%, unfortunately, there was a high incidence of pneumothorax [20]. On the other hand, combined with VBN, the positive rate of EBUS TBB on PPLs less than 2.0 cm in diameter could reach 44% [21]. Therefore, clinicians were concerned about how to increase the diagnostic yield of bronchoscopy, reduce the complications and shorten the procedure time.
Particularly, ROSE has revealled clinical importance in the diagnosis of lung tumors, lung nodules, mediastinum disease and other diseases [22]. Compared with the controls [23], ROSE could reduce the number of unnecessary punctures by 33%, meanwhile, it could make 68% of patients succeed in one puncture during TBNA. Although there were increasing researches on ROSE with the widespread application of TBNA, unfortunately, reports showed the controversial results [7,24,25]. Nakajima and his colleagues were in favor of our ndings that most patients with suspected lung cancer could be diagnosed by lung biopsy pathology in combination with ROSE, and the consistency between ROSE and the nal pathological diagnosis was 94.3% [24]. Conversely, Gri n et al. found that ROSE did not increase the positive rate of EBUS-TBNA, and it also did not reduce the number of punctures. Besides, ROSE increased expense, labor and time waste [25]. Also, there was no signi cant difference in diagnostic sensitivity and accuracy between ROSE group and the controls [7]. And it is now accepted that that ROSE could not increase the diagnostic yield of TBNA or EBUS-TBNA for skilled operators [26].
To the best of our knowledge, limited data are available concerning the application of ROSE during R-EBUS TBB in the diagnosis of PPLs [12][13][14]27]. A prospective RCT enrolling 152 patients with PPLs suggested that ROSE could improve the diagnostic yield and shorten the operation time. Meanwhile, no severe procedure related complications were observed, such as pneumothorax and hemorrhage [12]. The study was clearly in favour of our results, however, we should note that the researchers did not utilize brushing in the R-EBUS procedure, and the difference in the incidence of hemorrhage between both groups was signicant. In agreement with previous studies, we found that the number of unsatisfactory specimens in ROSE group decreased [14,17], then the diagnostic yield of PPLs based on ROSE expecially malignant tumors was improved [21]. However, the study conducted by Steinfort did not include control group, so the limitation may lead to some unnecessary bias [14]. Restricted to the sample size [12] and the cohort design [13,14], so the conclusions should be discussed with caution.
Our ndings demonstrated that ROSE could quickly evaluate whether the samples obtained are satisfactory, form a preliminary diagnosis in real time. Based on the good consistency beween ROSE and nal diagnosis, the positive result supported termination of procedure without further sampling. It is obvious that EBUS combined with ROSE can reduce the operation time [12,27]. And the improvement of the e ciency of bronchoscopy by ROSE can reduce the adverse physiological effects to some degree [28] and incidence of second biopsy. Futhermore, ROSE could show several neutrophils or macrophages or tumor cells, then bacterial culture or detailed biological characterization would be recommended [29]. In disagreement with previous study [12], Our data reported no signi cant difference in the operation time. But the diagnosis time was signi cantly shorter in ROSE group than the controls, and the. Importantly, we rstly reported the signi cantly decreased expense based on ROSE in combination with EBUS biopsy. And we speculated that the reduction of expense during diagnosis would be responsible for the reduced cases including severe complications and the chance of second biopsy, which contributed to public health resources greatly [30].
The majority of data suggested that some factors may affect the diagnostic yield of PPLs, such as the size of lesions and the location of lesions and so on [31]. Particularly, we were supported that ROSE can signi cantly improve the diagnostic yield of the lesions in diameter ≤ 2.0 cm [12]. And it may be attributed to the technical focus and tracheal structure reasons [31], which highlighted the clinical importance of this method in small lesions. On the other hand, supported by Steinfort [14], we observed no association between lesion size and ROSE outcome. Indeed, it should be noted that it was a cohort study and size of lesions uctuated greatly. Besides, our study was in accordance with previous researches that the diagnostic yield of upper lobe was relatively low [32,33]. We hypothesized that the bending angle of the upper lobe branch is too large, and the ultrasound probe is unable to stick to the focus and the sampling tool cannot extend to the distant focus according to the path of the probe. Under the guidance of ROSE, the extension path of biopsy forceps was modi ed to reach the lesion accurately, leading to an improvement of diagnostic yield in di cult cases.
There are still some limitations in our ndings. Firstly, although our study was a prospective RCT, it was a single-center design, the conclusion need to be interpreted with caution. Secondly, The sample size of our study was relatively small, so it needs to be expanded to reduce unnecessary bias in the future. Thirdly, we applied the Diff-Quik staining in our study, which is a modi cation of the Wright-Giemsa stain, whereas other researches used modi ed Shorr stain for slide preparations [34]. Different staining methods were reported to be associated with varying sensitivity [35]. We may need a optimal staining method for ROSE in future studies. Fourthly, it may be a little different in every patients' examination and treatment before diagnosis, which led to some bias in expense. In spite of this, it could still re ect every patient's cost during diagnosis to a certain degree.

Conclusions
In combination with ROSE, EBUS could increase the diagnostic yield of PPLs, shorten the diagnosis time, leading to a reduction of expense during diagnosis. ROSE would be of importance in diagnosing PPLs and medical burden.