Solitary Lung Nodule: CT-Guided Transthoracic Biopsy vs Transbronchial Biopsy With Endobronchial Ultrasound and Flexible Bronchoscope, a Meta-Analysis of Randomized Controlled Trials

Transbronchial lung biopsy with radial endobronchial ultrasound (rEBUS-TBB) and Computed tomography (CT) scan-guided transthoracic biopsy (CT-TTB) are commonly used to investigate peripheral lung nodules but high-quality data are still not clear about the diagnostic and safety profile comparison of these two modalities. We included all randomized controlled trials (RCT) comparing rEBUS-TBB with a flexible bronchoscope and CT-TTB for solitary lung nodules. Two reviewers extracted data independently on diagnostic performance and complication rates. 170 studies were screened, 4 RCT with a total of 325 patients were included. CT-TTB had a higher diagnostic yield than rEBUS-TBB (83.45% vs 68.82%, risk difference − 0.15, 95% CI, [− 0.24, − 0.05]), especially for lesion size 1–2 cm (83% vs 50%, risk difference − 0.33, 95% CI, [− 0.51, − 0.14]). For malignant diseases, rEBUS-TBB had a diagnostic yield of 75.75% vs 87.7% of CT-TTB. rEBUS-TBB had a significant better safety profile with lower risks of pneumothorax (2.87% vs 21.43%, OR = 0.12, 95% CI [0.05–0.32]) and combined outcomes of hospital admission, hemorrhage, and pneumothorax (8.62% vs 31.81%, OR 0.21, 95% CI, [0.11–0.40]). Factors increasing diagnostic yield of rEBUS were lesion size and localization of the probe but not the distance to the chest wall and hilum. CT-TTB had a higher diagnostic yield than rEBUS-TBB in diagnosing peripheral lung nodules, particularly for lesions from 1 to 2 cm. However, rEBUS-TBB was significantly safer with five to eight times less risk of pneumothorax and composite complications of hospital admission, hemorrhage, and pneumothorax. The results of this study only apply to flexible bronchoscopy with radial ebus without navigational technologies. More data are needed for a comparison between CT-TTB with rEBUS-TBB combined with advanced navigational modalities.


Introduction
Lung cancer continues to pose a significant burden on healthcare utilization and overall mortality. As of 2021, lung cancer remains the leading cause of death from cancer in the USA [1]. The results of the National Lung Screening Trial showed a mortality benefit of 14-20% by screening high-risk patients with low-dose Computed Tomography (CT) scan [2]. A solitary pulmonary nodule is defined as a discrete lesion measuring less than 3 cm noted in the lung parenchyma [3]. The risk of malignancy associated with such nodules varies by nodule characteristics and patients' risk factors [4]. Diagnosis can be established by tissue sampling by different modalities. Transbronchial biopsies with flexible bronchoscope guided by radial endobronchial ultrasound (rEBUS) and transthoracic needle aspiration with CT scan guidance are the two most common modalities widely utilized to biopsy lung nodules [4,5].
Choosing bronchoscopy with radial EBUS-guided transbronchial biopsy (rEBUS-TBB) or CT-guided transthoracic lung biopsy (CT-TTB) has been driven by clinical judgements based on the nodule characteristics and patient profile. Observational studies demonstrated that transthoracic needle 1 3 biopsy had the diagnostic accuracy of 90-93% [6,7]; however, complication rates of pneumothorax and bleeding were reported in more than 20% of patients [6]. Flexible bronchoscopy with radial endobronchial ultrasound guidance was reported in observational studies as a safer approach with a complication rate of only 2% and a diagnostic yield of 80% − 87.5% [8]. There have been multiple observational studies and randomized control trials comparing the transbronchial biopsy with radial EBUS vs transthoracic needle biopsy. Available high-quality randomized controlled trials demonstrated various results regarding the diagnostic value comparison between the two modalities.
This meta-analysis of the available randomized control trials aimed to assess the diagnostic yield of CT-guided transthoracic lung biopsy vs radial EBUS-guided transbronchial lung biopsy with the flexible bronchoscope and therefore helped to identify factors that would assist clinicians to select patients for these procedures.

Method
We used the following electronic databases: Pubmed (1970 to present), Scopus (1970 to present), Clinicaltrials.gov (1970 to present), and Virtual health library (1970 to present) to identify relevant articles in March 2021. Search terms were [(lung nodule) or (peripheral lung lesion)] and [(radial EBUS) or (rebus) or (radial endobronchial ultrasound) or (bronchoscopy)] and [(ct guided biopsy) or (computed tomography) or (ct fluoroscopy) or (CT cone beam) or (ct scan) or (percutaneous biopsy) or (transthoracic biopsy)] and [(randomized controlled trial) or (trial)]. The study protocol followed the recommendations of the Preferred Reporting item for Systematic Review and Meta-analysis (PRISMA) statement. A 2-level literature search was performed. A secondary search included manual scrutiny of the reference list of relevant articles received in the initial search after importing all searched studies from the databases into Endnote. English language restriction was applied. This review has not been registered.

Study Selection and Data Collection
We used the following inclusion criteria: (1) randomized controlled trial, (2) comparing CT-guided biopsy with radial EBUS performances in patients who had a single peripheral lung nodule or lesion, and (3) outcomes including diagnostic rate or diagnostic yield. Studies were excluded if they were (i) non-randomized controlled trials, (ii) conference posters, and (iii) did not measure mortality outcomes. Any difference in opinion about the eligibility was solved by discussion and consensus.

Full-Text Screening and Data Extraction
Two reviewers extracted the data independently into a spreadsheet. The data sought included study characteristics, including title, authors, year of publication, study location, number of patients, age, and important exclusion criteria.

Endpoints
Primary endpoint: diagnostic rates and diagnostic sensitivity.
Secondary endpoints: complication rates, factors influencing the diagnostic yield of radial EBUS.

Risk of Bias and Evidence Profile
We used the Cochrane Risk of Bias tool for a randomized control trial to assess the bias profile of individual trials. The tool included 7 domains: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other possible bias. Risks were classified as low, high risk, or unclear. Publication bias was analyzed using Egger's regression test and funnel plot. The quality of the evidence was assessed for the primary endpoint in agreement with the Grading of Recommendations Assessment Development and Evaluation system.

Statistical Analysis
Statistical analyses were performed using the Review Manager 5.3 software (Cochrane Collaborative Oxford, UK) and MAVIS 1.1.1 R package. Diagnostic yield (or diagnostic rate) was defined as the rate of diagnostic procedure in all procedures performed. Diagnostic sensitivity of a certain disease was calculated by the ratio between the positive diagnostic procedure for the disease and the total number of disease. Individual study relative risks (RRs) and 95% CI were calculated for dichotomous outcomes and mean differences (MDs) with 95% CI were estimated for continuous outcomes. Summary estimates of RRs were determined using the DerSimonian and Laird random effects models. The variability across studies not by chance was tested using the I 2 statistics and p > 0.05 was considered not statistically significant. Inconsistency across the studies were classified as low if 25% < I2 ≤ 50%, moderate if 50% < I2 ≤ 75%, and high if I2 > 75. Publication bias was analyzed as previously described by the funnel plot and regression test for funnel plot asymmetry. A p value of < 0.05 was considered a statistically significant publication bias.

Literature Search and Study Characteristics
Using the search term mentioned above, we found 170 studies. Further abstract screening found 5 case reports, unrelated studies, and 32 review articles. 4 studies met the inclusion criteria with the characteristics listed below in Table 1 [9][10][11][12]. There were a total of 325 patients included. Baseline demographic characteristics of age, nodule size, percentage of lower lobe location, and the malignancy diagnosis rate were not significantly different between the rEBUS-TBB and CT-TTB groups (67.68 ± 12.09 vs 61. 11

General Diagnostic Yield
Diagnostic biopsy definition was described in details in Table 1. Briefly, all studies defined a biopsy as diagnostic when a malignant pathology or a defined benign pathology was found (such as hamartoma and fungal infection). A non-diagnostic biopsy was required to have an additional biopsy either by surgery, CT-TTB or rEBUS-TBB. A true benign diagnosis was demonstrated if there was radiologic stability or improvement over 6-12 months (except there was no surveillance in study by Gupta et al.). The study by Gupta et al. only had 2 benign cases of tuberculosis. Figure 1 shows diagnostic comparisons between CT-TTB vs rEBUS-TBB. The CT-TTB group had a higher diagnostic rate than radial EBUS-TBB (83.45% vs 68.82%, risk difference − 0.15, 95% CI, [− 0.24, − 0.05]).
There was low heterogeneity in all the studies for all above outcome comparisons (I 2 = 0%).

Radial EBUS: Factors Determine Diagnostic Yield
Larger nodule size and probe localized within lesion increased the yield of radial EBUS. Two studies of 57 patients showed diagnostic group had larger nodule size compared to non-diagnostic group (mean difference 0.77, 95% CI [0.11, 1.42]) and higher rate of probe localized within lesion (OR, 18.42 95% CI, 1.80-188).
Other factors such as distance from the hilum, from the chest wall, and probe localized adjacent to lesion did not impact the diagnostic yield of rEBUS-TBB (Fig. 2).
There were moderate and high heterogeneity in the studies for the outcome comparison of location of the probe within or adjacent to the tumor.

Complications
Most common reported complications were pneumothorax and hemorrhage (either bleeding in the lesion or hemoptysis  (Fig. 3). There was no statistically significant difference in bleeding and hemoptysis rate between the two groups (5.7% vs 9%, p = 0.45).

Publication Bias
We carried out Egger's regression test and funnel plots to assess the presence of publication bias. Funnel plots showed no clear evidence of asymmetry and Egger's regression test indicated no evidence of publication bias (data not shown).

Discussion
To our knowledge, this study is the first meta-analysis of all available randomized trials to compare transbronchial biopsy with flexible bronchoscope and rEBUS vs transthoracic lung biopsy under CT guidance. CT-guided transthoracic lung biopsy had a higher diagnostic yield than bronchoscopy with radial EBUS in diagnosing peripheral lung nodule, particularly for lesion from 1 to 2 cm. The transbronchial approach with bronchoscopy and radial EBUS was significantly safer with five times to eight times less risk of pneumothorax and composite outcome of hospital admission, hemorrhage, and pneumothorax. Our meta-analysis also highlighted a few factors associated with the increased diagnostic yield of radial EBUS, including larger nodule sizes and probe location within lesions. However, the results of this meta-analysis cannot be generalized to all advanced navigational bronchoscopy technologies with radial ebus because there has been no randomized control trial to date comparing advanced navigational bronchoscopy and CT-TTB. All included randomized controlled trials were comparable with respect to the equipment used for radial EBUS group. In the rEBUS-TBB group, flexible bronchoscopes without electromagnetic guidance were utilized in all the studies. On the other hand, the type of CT scan varied in all the studies for CT-TTB group. Despite these variations, the overall diagnostic yield of CT-TTB was higher than radial EBUS (83.45% vs 68.82%). Non-diagnostic lesions required a repeat biopsy in all the studies. Benign lesions were followed up for radiographic stability over 6-12 months (except for the study by Gupta et al. in which there were only 2 benign diagnosis of tuberculosis). This is in line with the diagnostic workflow of other bronchoscopy studies [13,14]. As the result, the false-negative rate or missing a malignancy diagnosis was considered low.
The diagnostic yield of radial EBUS was reported to be comparable to that of CT-guided biopsy of peripheral lung nodule by Gupta  Overall complication rates were higher in CT-guided biopsy group when compared with radial EBUS group (31.81% vs 8.62%) with the majority being pneumothorax (21.43% vs 2.87%) followed closely by chest pain (18.46%vs 4.44%). Our findings of higher diagnostic yield but higher complication rate with CT-guided transthoracic lung biopsy are aligned with findings of prior meta-analysis of observational studies by Han et al. [8]. There are advantages to radial EBUS over CT-guided biopsy especially when complication rate is considered. Selection of the right patient for radial EBUS is the key.  expected when analyzing studies across wide geographic and temporal distribution. Overall bias was considered high in all these studies primarily due to high risks of performance bias.
The results of this study could only be applied to flexible bronchoscopy with the outer diameter more than 4.9 mm and radial ebus without advanced navigational bronchoscopy technologies. In the past decade, there have been multiple developments, such as robotic bronchoscopy, electromagnetic navigational bronchoscopy, as well as imaging modalities, such as 3D fluoroscopy and cone-beam CT scan. With these new developments, biopsy of smaller nodules less than 1-2 cm in challenging positions has become increasingly feasible [15,16]. It is also important to note that the diagnostic yield of ultrathin bronchoscope (outer diameter of 3 mm, working channel of 1.7 mm) with radial ebus (1.4 mm) was proven to be higher than the thin bronchoscope (outer diameter of 4 mm) with radial ebus [17]. More studies are needed in future to compare these new modalities to the traditional transthoracic CT-guided biopsy. Another fact to know is most of the studies in this meta-analysis used moderate sedations. The complications from general anesthesia which are used extensively for advanced diagnostic navigational bronchoscopy will also need to be evaluated in future studies. However, the results of this study are still helpful, especially for areas with limited resources where there are no available advanced technologies, such as robotic or electromagnetic navigation. To select the appropriate biopsy modality for suspected lung cancer, clinicians should also know the important advantage of the transbronchial biopsy approach which is the possibility to incorporate mediastinal staging procedure with linear endobronchial ultrasound in the same setting. This advantage could help to reduce the wait time to both diagnosis and staging for the patients.

Conclusion
In summary, our meta-analysis demonstrated that CT-TTB had a higher diagnostic yield than rEBUS-TBB in diagnosing peripheral lung nodules, particularly for lesions from 1 to 2 cm. However, rEBUS-TBB was significantly safer with five to eight times less risk of pneumothorax and composite complications of hospital admission, hemorrhage, and pneumothorax. Underlying structural lung disease such as emphysema should also be factored while assessing the risks of complications from procedure. With the advent of navigational bronchoscopy and ultrathin bronchoscopes, randomized control trials comparing the diagnostic yield of CTguided biopsy vs navigational bronchoscopy would further benefit the clinicians in planning the diagnostic workup for patients with peripheral lung nodule.