Abstract
Background
Despite the many advances in peripheral bronchoscopy, its diagnostic yield remains suboptimal. With the use of cone-beam CT imaging we have found atelectasis mimicking lung tumors or obscuring them when using radial-probe endobronchial ultrasound (RP-EBUS), but its incidence remains unknown.
Research Question
What are the incidence, anatomic location, and risk factors for developing atelectasis during bronchoscopy under general anesthesia?
Study Design and Methods
We performed a prospective observational study in which patients undergoing peripheral bronchoscopy under general anesthesia were subject to an atelectasis survey carried out by RP-EBUS under fluoroscopic guidance. The following dependent segments were evaluated: right bronchus 2 (RB2), RB6, RB9, and RB10; and left bronchus 2 (LB2), LB6, LB9, and LB10. Images were categorized either as aerated lung (“snowstorm” pattern) or as having a nonaerated/atelectatic pattern. Categorization was performed by three independent readers.
Results
Fifty-seven patients were enrolled. The overall intraclass correlation agreement among readers was 0.82 (95% CI, 0.71-0.89). Median time from anesthesia induction to atelectasis survey was 33 min (range, 3-94 min). Fifty-one patients (89%; 95% CI, 78%-96%) had atelectasis in at least one of the eight evaluated segments, 45 patients (79%) had atelectasis in at least three, 41 patients (72%) had atelectasis in at least four, 33 patients (58%) had atelectasis in at least five, and 18 patients (32%) had atelectasis in at least six segments. Right and left B6, B9, and B10 segments showed atelectasis in > 50% of patients. BMI and time to atelectasis survey were associated with increased odds of having more atelectatic segments (BMI: OR, 1.13 per unit change; 95% CI, 1.034-1.235; P = .007; time to survey: OR, 1.064 per minute; 95% CI, 1.025-1.105; P = .001).
Interpretation
The incidence of atelectasis developing during bronchoscopy under general anesthesia in dependent lung zones is high, and the number of atelectatic segments is greater with higher BMI and with longer time under anesthesia.
Clinical Trial Registration
ClinicalTrials.gov; No.: NCT03523689; URL: www.clinicaltrials.gov.
Key Words: atelectasis, bronchoscopy, general anesthesia, radial endobronchial ultrasound
Abbreviations: CBCT, cone-beam CT; CP-EBUS, convex-probe endobronchial ultrasound; LMA, laryngeal mask airway; RP-EBUS, radial-probe endobronchial ultrasound
FOR EDITORIAL COMMENT, SEE PAGE 2268
In the past two decades, we have witnessed the development of multiple bronchoscopy techniques focused on the diagnosis of peripheral lung nodules. Although peripheral bronchoscopy has been established as a safe procedure, its diagnostic yield seems to have plateaued at about 70%.1, 2, 3, 4, 5 Technologies such as ultrathin bronchoscopy, radial-probe endobronchial ultrasound (RP-EBUS), electromagnetic and nonelectromagnetic navigation, and more recently robotic bronchoscopy continue to fail to deliver a higher diagnostic yield. Although these technologies seemed to have improved our ability to reach the target—navigational yield—there continues to be a gap between navigational yield and diagnostic yield.6 In the absence of an intrabronchoscopy CT guidance, the only available real-time method to corroborate that a lesion has been reached (navigational yield) is RP-EBUS. However, a study by our group on cone-beam CT (CBCT)-guided bronchoscopy showed that patients commonly develop atelectasis during bronchoscopy under general anesthesia.7 Because of its increased density compared with aerated lung, atelectatic lung tissue can potentially give a false positive RP-EBUS finding—mimicking a lung lesion—and lead to inadvertent biopsy of normal lung tissue. Even when we recognize these images as atelectasis, they can still completely obscure our targets and prevent us from obtaining a diagnosis.
The inaccuracy of common navigation techniques, such as electromagnetic navigational bronchoscopy, has been blamed on the concept of “CT-body divergence.” This phenomenon can be described as the discrepancy between the prebronchoscopy chest CT imaging used for navigation, and the true shape, volume of the lungs, and position of the target found during bronchoscopy. It is hypothesized that atelectasis plays a major role in CT-body divergence, in addition to the excursion of lung targets with respiratory movements.6, 7, 8
Given the high clinical impact of the development of atelectasis during peripheral bronchoscopy described above, we designed a prospective study to describe the incidence of intraprocedural atelectasis during bronchoscopy as detected by RP-EBUS, to determine their specific anatomic location, and to describe patient- and procedure-related risk factors for the development of atelectasis.
Materials and Methods
Study Setting and Subjects
The study was performed at the University of Texas MD Anderson Cancer Center, and it was approved by its institutional review board (2018-0123). Adult patients referred to our pulmonary department for diagnosis of peripheral lung nodules, and who were considered for peripheral bronchoscopy with RP-EBUS, were considered for this trial. All patients provided written informed consent. Only patients who had undergone recent chest CT imaging (within 4 weeks of bronchoscopy) were considered. Patients with evidence on CT imaging of lung consolidation, interstitial changes, or lung tumors (≥ 2 cm in diameter in dependent areas [our target study bronchial segments as described below]); pregnant patients; patients with known pleural effusions or ascites; and patients with diaphragmatic paralysis were excluded. Patients who were found to have secretions during bronchoscopy in the target study airways that were considered to be more than mild and a possible cause of atelectasis by the bronchoscopist were also excluded.
Study Design and Procedures
This was a prospective observational study in which patients undergoing peripheral bronchoscopy (with or without mediastinal staging with convex-probe EBUS [CP-EBUS]) were subject to an atelectasis survey carried out by RP-EBUS under fluoroscopic guidance. Bronchoscopy was performed in accordance with our standard of care under general anesthesia (total IV anesthesia) with ventilation being provided through a laryngeal airway mask (LMA) and all patients receiving neuromuscular blocking agents. Once the diagnostic and staging (if applicable) procedure was over, the time was recorded, and the atelectasis survey was carried out. When a tumor was located in one of the eight study target bronchial segments, the atelectasis survey was performed before biopsy of the tumor to prevent bleeding or wedging of the scope, which might induce atelectasis. Using the RP-EBUS (with 4-cm depth scale), the following eight study target segments in dependent areas of the lungs were systematically evaluated: RB2, RB6, RB9, and RB10; LB2, LB6, LB9, and LB10. For each segment, two subsegments (a and b) were evaluated. The radial probe was advanced carefully in each subsegment under fluoroscopic guidance until resistance was felt while monitoring the ultrasound images. The probe was then withdrawn approximately 3 cm while monitoring ultrasound images. If only aerated lung was observed, an image at any point on the probe trajectory was recorded. When an abnormal image (any hypoechoic pattern other than a vessel, of at least 2 cm in diameter) was seen in at least 2 cm of the probe trajectory, this nonaerated pattern was recorded. Ultrasound pictures were categorized either as aerated lung (“snowstorm” pattern) or as having a nonaerated/atelectatic pattern (Fig 1). The initial categorization was performed by the bronchoscopist, and images were then reviewed postprocedure by another two independent reviewers who were blinded to prior categorization, CT images, and medical records to ensure validity of the interpretation. All reviewers were highly experienced in RP-EBUS. Additional collected data included basic demographic information, the time from induction of anesthesia (placement of LMA) to the capturing of images, ventilatory settings, and drugs used for anesthesia, among others. Tidal volume was recorded immediately before the survey for atelectasis by RP-EBUS, without a bronchoscope in the airways. Study data were collected and managed with REDCap (Research Electronic Data Capture) hosted at our institution.9
Figure 1.
Radial-probe endobronchial ultrasound patterns. A, Aerated lung (snowstorm). B-D, Various patterns of nonaerated lung or atelectasis: concentric “tumor-like” with clear borders (B); concentric irregular with poorly demarcated borders (C); eccentric (D). (All images taken with 4-cm depth scanning).
Atelectasis was considered present in any bronchial segment when at least one of the two subsegments (a and b) was categorized as such (nonaerated pattern on RP-EBUS) by at least two of the three independent readers. The primary objective was to estimate the proportion of patients identified as developing intraprocedural atelectasis. Secondary end points included the location of atelectasis, the number of atelectatic segments per patient, and the identification of patient and procedural characteristics that may predispose to the development of atelectasis.
Statistics
For each patient, a binary outcome variable indicating development of at least one atelectatic segment or not was determined. On the basis of the incidence of atelectasis described in our prior study on cone-beam CT scan-guided bronchoscopy, an anticipated incidence, the proportion of patients developing atelectasis detected by RP-EBUS, was 0.2.7 The sample of 57 patients was required to construct a 95% CI for the proportion of patients developing atelectasis with a margin of 10%, assuming an estimated proportion of 0.2 (PASS 2005; NCSS Statistical Software). The proportion of patients identified as developing intraprocedural atelectasis by RP-EBUS and its 95% CI were estimated. Locations of atelectasis were summarized by a frequency table, allowing for multiple locations in the same patient. The proportion of segments identified as developing atelectasis among all evaluated segments was calculated for each patient, and the distribution of the proportions was summarized as the mean, median, and range. The intraclass correlation agreement among the three readers of RP-EBUS images was assessed by Shrout and Fleiss intraclass correlations.10 The Fisher z transformation was used to construct a 95% CI for intraclass correlation. Intraclass correlation coefficients above 0.8 or 0.9 were regarded as good or excellent agreement. We compared patients with zero to three, four or five, and six to eight segments with atelectasis in terms of patient characteristics, using the Kruskal-Wallis test for continuous variables and the Fisher exact test or χ2 test for categorical variables. To identify patient characteristics associated with the atelectasis groups (zero to three, four or five, and six to eight segments with atelectasis), univariate ordinal logistic regression models were used. Multivariate ordinal logistic regression models included variables with a significant P value based on univariate models. The proportional odds assumption was checked and the Hosmer-Lemeshow test was performed to check the goodness of fit for logistic regression models. A P value < .05 was used to indicate statistical significance. SAS 9.4 (SAS Institute Inc) was used for data analysis.
Results
Fifty-seven patients were included in the study and 456 bronchial segments were surveyed for atelectasis by RP-EBUS. Baseline characteristics are depicted in Table 1. A total of 904 RP-EBUS images were prospectively recorded and assessed by our three independent reviewers (two images per segment, eight segments per patient, eight pictures missing due to technical errors). The overall intraclass correlation agreement among readers was 0.82 (95% CI, 0.71-0.89) for all segments, 0.57 (95% CI, 0.37-0.72) for upper lobe segments, and 0.87 (95% CI, 0.79-0.92) for lower lobe segments. The median time from anesthesia induction to atelectasis survey was 33 min (range, 3-94 min). All subjects underwent diagnostic peripheral bronchoscopy and 36 of them (63%) also had mediastinal staging. Six patients with tumors within a study bronchial segment had the atelectasis survey performed before diagnostic bronchoscopy. Six of 456 surveyed bronchial segments had a tumor located within that segment (1.3%), and none of these six segments were read as atelectatic. Fifty-one patients (89%; 95% CI, 78%-96%) were found to have atelectasis in at least one of the eight evaluated segments, 45 patients (79%) had atelectasis in at least three, 41 patients (72%) had atelectasis in at least four, 33 patients (58%) had atelectasis in at least five, and 18 patients (32%) had atelectasis in at least six of the eight evaluated segments. LB10 (79%) and LB6 (70%) were the most common locations for the development of intraprocedural atelectasis. Right and left B6, B9, and B10 segments showed atelectasis in > 50% of the patients (Fig 2). The mean proportion of evaluated segments per patient that was identified as developing atelectasis was 54% (interquartile range, 37.5%-75%). Atelectasis was not evident by fluoroscopy in any of the cases. A comparison of patient characteristics when grouped on the basis of the number of atelectatic segments (zero to three, four or five, six to eight) is summarized in Table 2. For groups with a greater number of atelectatic segments, BMI was higher and time from artificial airway to survey was longer. Table 3 summarizes our univariate ordinal logistic regression analysis results. Higher BMI (as a continuous variable and a dichotomized variable) and longer time from artificial airway to survey (as a continuous variable and a dichotomized variable) showed significant associations with increased odds of developing a greater number of segments with atelectasis. Table 4 summarizes a multivariable ordinal logistic regression analysis result. By increasing 1 unit in BMI, the odds of having six to eight atelectatic segments vs the odds of having zero to five atelectatic segments, and the odds of having four to eight atelectatic segments vs the odds of having zero to three atelectatic segments, are increased by 1.13 (95% CI, 1.034-1.235). By increasing 1 min from artificial airway to survey, the odds of having six to eight atelectatic segments vs the odds of having zero to five atelectatic segments, and the odds of having four to eight atelectatic segments vs the odds of having zero to three atelectatic segments, are increased by 1.064 (95% CI, 1.025-1.105). The proportional odds assumption was not violated (P = .8756) and the Hosmer-Lemeshow test showed a good model fit (P = .5814). A person with higher BMI is more likely to have a higher number of atelectatic segments. Similarly, a person with longer time from artificial airway to survey is more likely to have a higher number of atelectatic segments.
Table 1.
Baseline Patient and Procedure Characteristics
| Characteristic | Value |
|---|---|
| Subjects, No. | 57 |
| Age, median (range), y | 71 (41-90) |
| Sex | |
| Female, No. (%) | 30 (53) |
| Male, No. (%) | 27 (47) |
| BMI, median (range), kg/m2 | 26.4 (17.51-49.03) |
| Indication for bronchoscopy, No. (%) | |
| Diagnostic peripheral bronchoscopy | 57 (100) |
| Mediastinal staging | 36 (63) |
| Tidal volume, mean ± SD, mL | 434.45 ± 80.89 |
| Tidal volume per kg ideal body weight, mean ± SD, mL/kg | 9.07 ± 1.45 |
| Fio2, median (range), % | 100 (50-100) |
| PEEP, median (range), cm H2O | 0 (0-7) |
| Time to atelectasis survey, median (range), min | 33 (3-94) |
PEEP = positive end-expiratory pressure.
Figure 2.
Incidence of atelectasis in dependent lung segments (posterior lung views). LB = left bronchus; RB = right bronchus.
Table 2.
Characteristics of Patients With Various Numbers of Atelectatic Segments
| Characteristic | No. of Segments With Atelectasis per Patient |
P Value | ||
|---|---|---|---|---|
| 0-3 | 4 or 5 | 6-8 | ||
| BMI, mean ± SD, kg/m2 | 25.85 ± 6.44 | 28.41 ± 7.03 | 30.41 ± 5.45 | .0554 |
| Tidal volume, mean ± SD, mL | 430.33 ± 86.86 | 431.82 ± 86 | 441.11 ± 73.24 | .7662 |
| Tidal volume per kg ideal body weight, mean ± SD, mL/kg | 8.98 ± 1.53 | 8.92 ± 1.6 | 9.33 ± 1.22 | .3426 |
| Fio2, mean ± SD, % | 94.31 ± 13.35 | 97.74 ± 6.3 | 98.78 ± 2.92 | .4434 |
| PEEP, mean ± SD, cm H2O | 1.5 ± 2.31 | 0.78 ± 1.28 | 1.17 ± 1.58 | .7020 |
| Time to atelectasis survey, mean ± SD, min | 30.5 ± 15.54 | 36.17 ± 13.94 | 47.22 ± 19.36 | .0095 |
| BMI, No. (%), kg/m2 | .0201 | |||
| < 25 | 10 (62.5) | 11 (47.8) | 3 (16.7) | |
| ≥ 25 | 6 (37.5) | 12 (52.2) | 15 (83.3) | |
| Tidal volume, No. (%), mL | .3570 | |||
| < 400 | 6 (37.5) | 10 (43.5) | 4 (22.2) | |
| ≥ 400 | 10 (62.5) | 13 (56.5) | 14 (77.8) | |
| Fio2, No. (%) | .5903 | |||
| < 100 | 6 (37.5) | 6 (26.1) | 4 (22.2) | |
| 100 | 10 (62.5) | 17 (73.9) | 14 (77.8) | |
| PEEP, No. (%), cm H2O | .8150 | |||
| 0 | 10 (62.5) | 15 (65.2) | 10 (55.6) | |
| ≥ 1 | 6 (37.5) | 8 (34.8) | 8 (44.4) | |
| Time to atelectasis survey, No. (%) | .0820 | |||
| < 33 min | 13 (81.3) | 15 (65.2) | 8 (44.4) | |
| ≥ 33 min | 3 (18.8) | 8 (34.8) | 10 (55.6) | |
See Table 1 legend for expansion of abbreviation.
Table 3.
Univariate Ordinal Logistic Regression Model
| Covariate | OR | 95% CI | P Value |
|---|---|---|---|
| BMI, 1-unit change | 1.083 | 1.000-1.172 | .0488 |
| Tidal volume per kg IBW, 1-unit change, mL | 1.131 | 0.803-1.594 | .4799 |
| Fio2, 1-unit change | 1.053 | 0.982-1.130 | .1482 |
| PEEP, 1-unit change, cm H2O | 0.920 | 0.693-1.223 | .5664 |
| Time to atelectasis survey, 1-unit change, min | 1.049 | 1.015-1.084 | .0043 |
| BMI group | |||
| < 25 | 1.000 | … | … |
| ≥ 25 | 4.167 | 1.467-11.837 | .0074 |
| Tidal volume group | |||
| < 400 mL | 1.000 | … | … |
| ≥ 400 mL | 1.642 | 0.597-4.517 | .3369 |
| Tidal volume per kg IBW | |||
| < 8.77 mL/kg | 1.000 | … | … |
| ≥ 8.77 mL/kg | 0.914 | 0.343-2.435 | .8569 |
| Fio2 group | |||
| < 100% | 1.000 | … | … |
| 100% | 1.716 | 0.585-5.033 | .3252 |
| PEEP group, cm H2O | |||
| 0 | 1.000 | … | … |
| ≥ 1 | 1.251 | 0.467-3.354 | .6565 |
| Time to atelectasis survey group | |||
| < 33 min | 1.00 | … | … |
| ≥ 33 min | 3.255 | 1.145-9.256 | .0269 |
IBW = ideal body weight. See Table 1 legend for expansion of other abbreviation.
Table 4.
Multivariate Ordinal Logistic Regression Model
| Covariate | ORa | 95% CI | P Value |
|---|---|---|---|
| BMI, 1-unit change | 1.130 | 1.034-1.235 | .0071 |
| Time to atelectasis survey, 1-unit change, min | 1.064 | 1.025-1.105 | .0011 |
Odds of having a higher vs a lower number of atelectatic segments (eg, number of atelectatic segments of six to eight vs zero to five or four to eight vs zero to three).
Discussion
Atelectasis seems to play a detrimental role in our practice of peripheral bronchoscopy. Atelectases can obscure or mimic targets with RP-EBUS, and they may be largely responsible for the CT-body divergence that widens the margin of error of most navigational techniques. To the best of our knowledge, our current study is the first to describe the incidence of intraprocedural atelectasis during bronchoscopy under general anesthesia in each specific dependent lung segment. Our finding of an incidence of > 50% of atelectasis in all dependent lower lobe segments after 30 min of anesthesia, although not surprising, is of high relevance and it is bound to affect our peripheral bronchoscopy practice.
Atelectasis has been described in up to 90% of anesthetized patients undergoing various surgical procedures independent of age, sex, or anesthetics used.11 This phenomenon has not been described in the literature in the setting of bronchoscopy, likely because most bronchoscopies have traditionally been performed under moderate/conscious sedation. With the exponential growth of the field of bronchoscopy in the past two decades, pulmonologists now play a major role in the diagnosis and staging of lung cancer, and they perform much more complex and prolonged procedures. We believe that the increasing length of these procedures has led bronchoscopists in many centers to perform these under general anesthesia for multiple reasons: patient comfort, concomitant use of on-site cytology, greater time to educate trainees, and ease of procedure, among others. In addition to general anesthesia, atelectasis can occur during peripheral bronchoscopy for other reasons such as wedging of the bronchoscope in a given bronchial segment and bleeding/clotting.
As described earlier, the intraprocedural development of atelectasis primarily impacts peripheral bronchoscopy procedures. Whereas most peripheral bronchoscopy data report a navigational yield of about 90% or even higher, the diagnostic yield remains closer to 70%.2,4,5,8,12 This gap likely has more than one explanation. The inability of our current biopsy tools to obtain proper samples can be one. But more recently, with the aid of intrabronchoscopy use of CT imaging, we have discovered that our navigational yield may, indeed, have been overestimated. We have traditionally considered our navigational yield successful with either the finding of a positive RP-EBUS image, positive pathology, or less commonly (and less accurately) trusting the navigational software determination that we had reached the target. With > 50% of patients developing atelectasis in dependent areas in our study, we strongly believe that this phenomenon plays an important role in this gap. False positive RP-EBUS images have likely led many of us to take unnecessary and nondiagnostic biopsies of atelectatic lung parenchyma for many years. Biopsies showing blood, bronchial cells, and alveolar macrophages obtained from lesions that showed a concentric RP-EBUS image are possibly explained by this phenomenon. Although experienced bronchoscopists may potentially distinguish atelectasis from tumors and avoid these nondiagnostic samples, atelectasis can still obscure their targets, prevent them from localizing them, and prevent them from obtaining a diagnosis. A multicenter randomized controlled trial of standard fluoroscopy-guided bronchoscopy vs thin bronchoscopy with RP-EBUS by Tanner and coworkers2 reported one of the greatest gaps between navigation and diagnostic yield we have found in the literature. Of 179 patients who underwent RP-EBUS either because of randomization or subsequent crossover, 174 (97%) had ultrasound confirmation of lesion localization, with a concentric image seen in 113 (65%). Yet, diagnostic yield was 50% for concentric lesions and 31% for eccentric lesions.
Electromagnetic, nonelectromagnetic, and shape-sensing bronchoscopic navigational software have been developed in the past two decades. All of these rely on prebronchoscopy chest CT imaging for both planning and navigation phases. Great relevance was given to the respiratory phase-induced changes of the lung (volume and shape) and its impact on the navigational accuracy of these techniques. The concept of the moving target and its excursion throughout the respiratory phases has been well described.8,12 Adding to these barriers to accurate navigation, we now hypothesize that, when performed under general anesthesia, the development of atelectasis also plays an important role in CT-body divergence. Atelectasis can either shorten the airways, or pull them and the targets in different directions, away from their original location on the planning CT scan. Studies with intraprocedural chest CT imaging or three-dimensional fluoroscopy to better characterize these conformational changes and the true impact of atelectasis in this aspect are needed.
Our study reports that the longer the time to the atelectasis survey (the longer the time under general anesthesia), the greater the number of atelectatic lung segments. This is highly relevant in our bronchoscopy practice, particularly when dealing with patients with suspected lung cancer. In patient with lung lesions that are suggestive of lung cancer, or who have enlarged (> 1 cm in short axis) or PET-positive hilar or mediastinal lymph nodes, centrally located lung tumors, or lung masses (> 3 cm) CP-EBUS for nodal staging is indicated.13,14 When planning a combined CP-EBUS and peripheral bronchoscopy procedure, if rapid on-site cytology is available, CP-EBUS is typically performed before peripheral bronchoscopy. This way, if any of the lymph nodes are positive for malignancy, there is no need for peripheral bronchoscopy, which carries a greater rate of complications than CP-EBUS. Reported median durations of systematic staging with CP-EBUS (N3 to N2 to N1) range from 20 to 40 min, with many factors influencing this time (ie, on-site cytology, involvement of trainees).15, 16, 17, 18, 19 Evaluated at a median time of 33 min, well within the range of time necessary for a staging EBUS, atelectases were quite common in our study. Thus, carrying out CP-EBUS for staging before peripheral bronchoscopy can negatively influence the yield of the latter.
The incidence of atelectasis we describe in the current study is greater than the one we described in our prior pilot study of cone-beam CT-guided bronchoscopy.7 There are multiple possible explanations for this, with the most likely one being the fact that CBCT scans were typically performed within < 10 min, whereas our median survey with RP-EBUS was done at 33 min, and atelectases are more common with longer periods of general anesthesia. Also, we do not know which technique is more sensitive in detecting atelectasis. RP-EBUS can clearly detect parenchymal changes that cannot be evidenced by two-dimensional fluoroscopy, and it may indeed be able to detect changes that are not evident in CBCT scans, but this remains unknown.
RP-EBUS images characteristics of atelectatic lung have not been thoroughly described, and the use of RP-EBUS to detect them may be a potential limitation of our study. We believe that in the absence of previous parenchymal disease, tumors, secretions, or clots—all exclusion criteria in our study—there is no further explanation to the finding of new nonaerated areas of > 2 cm in diameter by RP-EBUS than the development of atelectasis. Concerned about the interpretation of RP-EBUS images, in addition to the individual performing bronchoscopy, we added two independent reviewers who were blinded to prior categorization, CT images, and medical records to ensure the validity of the interpretation. The overall intraclass correlation agreement among the three readers was very good (0.82), being higher (0.87) for lower lobe segments where the incidence of atelectasis was much higher, and lower (0.57) for the upper lobe segments where the incidence of atelectasis was lower, as expected. In our experience, after analyzing this vast number of images, the presence of normal-looking vasculature crossing a solid/semisolid RP-EBUS image is suggestive of atelectasis. However, this vascular pattern is not always present, and some solid images can be indistinguishable from a tumor (Fig 1B). Moreover, large tumors can also have visible vasculature. Tumors tend to have better defined edges that can help us differentiate them. Atelectases do not always have clear edges, and they tend to diffusely occupy the entire RP-EBUS screen. If the size of the solid RP-EBUS image is much larger than that of the target tumor, this would also indicate the presence of atelectasis.
Regarding limitations, our results may not be extrapolated to other centers where general anesthesia is provided with a different artificial airway or different mechanical ventilation parameters. We did not use a particular ventilator protocol, because we wanted to study the incidence of atelectasis with our current anesthesia practice. The vast majority of patients in our department undergo procedures with an LMA because it grants us easy access to upper lymph node stations during mediastinal staging, which is performed in most of these patients. We have also noticed that almost all of the patients in our department are ventilated with an Fio2 of 100% and with zero or minimal positive end-expiratory pressure, and we preferred doing our study under these same conditions, understanding that we would not be able to find an association of different Fio2 or positive end-expiratory pressure levels with the development of atelectasis. Another potential limitation of our study is the fact that the prebronchoscopy CT imaging could be up to 4 weeks old. Some patients may have developed new infiltrates or atelectasis in the interim, which may not have been related to the use of anesthesia or to bronchoscopy, although we believe this is unlikely. To our knowledge, by the time of bronchoscopy, none of these patients had any signs or symptoms that would have been associated with any such parenchymal opacities. The lack of consistency in the time to atelectasis survey among patients may be viewed as a limitation as well. Setting a specific time to make this atelectasis survey would have led to an interruption in the flow of the procedure that we decided to avoid. In addition, unless multiple surveys had been done at prespecified times, we would not have been able to gather information regarding the incidence of atelectasis in relation to the time spent under general anesthesia. Our median time ended up being 33 min, which is similar to the time it takes us to perform EBUS for systematic staging, and the time when the peripheral bronchoscopy (which could be affected by atelectasis) would typically start.
Conclusions
Our study demonstrates that the incidence of atelectasis developing during bronchoscopy under general anesthesia with LMA in dependent lung zones is high, and the number of atelectatic segments is greater with higher BMI and with longer periods under anesthesia. Atelectasis can create false positive images by RP-EBUS, obscure lung tumors, or cause CT-body divergence, all contributing to our inability to break through the seemingly capped diagnostic yield of peripheral bronchoscopy. Positive RP-EBUS images in dependent lower lobe segments need to be analyzed with extreme caution by bronchoscopists, and the fact that they could represent atelectasis should be kept in mind. Whether peripheral bronchoscopy should be performed before EBUS for mediastinal staging (in combined procedures), or whether specific ventilatory strategies can prevent this phenomenon, needs to be investigated.
Acknowledgments
Author contributions: R. F. C. is the guarantor of the content of the manuscript. He had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. A.-E. S. S., B. F. S., G. A. E., and R. F. C. contributed to the study design, bronchoscopy, image analysis, data collection and interpretation, and manuscript composition and revision. M. M. contributed to bronchoscopy, data collection and interpretation, and manuscript composition and revision. J. S. contributed to study design, statistical data analysis, and manuscript composition and revision. M. H. A. contributed to study design, data collection, data interpretation, and manuscript composition and revision. M. S., H. B. G., D. E. O., and C. A. J. contributed to study design, data interpretation, and manuscript composition and revision.
Financial/nonfinancial disclosures: The authors have reported to CHEST the following: R. F. C. has received research grants from Concordia, Siemens, Nanobiotix, and Olympus. He is a paid consultant for Olympus and Siemens. None declared (A.-E. S. S., B. F. S., G. A. E., J. S., M. M., M. S., M. H. A., H. B. G., D. E. O., C. A. J.).
Role of sponsors: The sponsor had no role in the design of the study, the collection and analysis of the data, or the preparation of the manuscript.
Other contributions: The authors thank Lee Taylor, Mehrnoosh Amirian, Justin Hair, Rodney Green, and William Newton for assistance with bronchoscopy and image recording. Without their invaluable assistance, this study would not have been possible.
Footnotes
FUNDING/SUPPORT: This research is supported in part by the National Institutes of Health [MD Anderson Cancer Center Support Grant CA016672].
References
- 1.Wang Memoli J.S., Nietert P.J., Silvestri G.A. Meta-analysis of guided bronchoscopy for the evaluation of the pulmonary nodule. Chest. 2012;142(2):385–393. doi: 10.1378/chest.11-1764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Tanner N.T., Yarmus L., Chen A. Standard bronchoscopy with fluoroscopy vs thin bronchoscopy and radial endobronchial ultrasound for biopsy of pulmonary lesions: a multicenter, prospective, randomized trial. Chest. 2018;154(5):1035–1043. doi: 10.1016/j.chest.2018.08.1026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sainz Zuñiga P.V., Vakil E., Molina S., Bassett R.L., Jr., Ost D.E. Sensitivity of radial endobronchial ultrasound-guided bronchoscopy for lung cancer in patients with peripheral pulmonary lesions: an updated meta-analysis. Chest. 2020;157(4):994–1011. doi: 10.1016/j.chest.2019.10.042. [DOI] [PubMed] [Google Scholar]
- 4.Khandhar S.J., Bowling M.R., Flandes J. Electromagnetic navigation bronchoscopy to access lung lesions in 1,000 subjects: first results of the prospective, multicenter NAVIGATE study. BMC Pulm Med. 2017;17(1):59. doi: 10.1186/s12890-017-0403-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Chaddha U., Kovacs S.P., Manley C. Robot-assisted bronchoscopy for pulmonary lesion diagnosis: results from the initial multicenter experience. BMC Pulm Med. 2019;19(1):243. doi: 10.1186/s12890-019-1010-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Casal R.F. Cone beam CT-guided bronchoscopy: here to stay? J Bronchology Interv Pulmonol. 2018;25(4):255–256. doi: 10.1097/LBR.0000000000000546. [DOI] [PubMed] [Google Scholar]
- 7.Casal R.F., Sarkiss M., Jones A.K. Cone beam computed tomography-guided thin/ultrathin bronchoscopy for diagnosis of peripheral lung nodules: a prospective pilot study. J Thorac Dis. 2018;10(12):6950–6959. doi: 10.21037/jtd.2018.11.21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chen A., Pastis N., Furukawa B., Silvestri G.A. The effect of respiratory motion on pulmonary nodule location during electromagnetic navigation bronchoscopy. Chest. 2015;147(5):1275–1281. doi: 10.1378/chest.14-1425. [DOI] [PubMed] [Google Scholar]
- 9.Harris P.A., Taylor R., Thielke R., Payne J., Gonzalez N., Conde J.G. Research electronic data capture (REDCap): a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377–381. doi: 10.1016/j.jbi.2008.08.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Shrout P.E., Fleiss J.L. Intraclass correlations: uses in assessing rater reliability. Psychol Bull. 1979;86(2):420–428. doi: 10.1037//0033-2909.86.2.420. [DOI] [PubMed] [Google Scholar]
- 11.Tusman G., Böhm S.H., Warner D.O., Sprung J. Atelectasis and perioperative pulmonary complications in high-risk patients. Curr Opin Anaesthesiol. 2012;25(1):1–10. doi: 10.1097/ACO.0b013e32834dd1eb. [DOI] [PubMed] [Google Scholar]
- 12.Furukawa B.S., Pastis N.J., Tanner N.T., Chen A., Silvestri G.A. Comparing pulmonary nodule location during electromagnetic bronchoscopy with predicted location on the basis of two virtual airway maps at different phases of respiration. Chest. 2018;153(1):181–186. doi: 10.1016/j.chest.2017.06.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Silvestri G.A., Gonzalez A.V., Jantz M.A. Methods for staging non-small cell lung cancer: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2013;143(5 suppl):e211S–e250S. doi: 10.1378/chest.12-2355. [DOI] [PubMed] [Google Scholar]
- 14.De Leyn P., Dooms C., Kuzdzal J. Revised ESTS guidelines for preoperative mediastinal lymph node staging for non-small-cell lung cancer. Eur J Cardiothorac Surg. 2014;45(5):787–798. doi: 10.1093/ejcts/ezu028. [DOI] [PubMed] [Google Scholar]
- 15.Casal R.F., Lazarus D.R., Kuhl K. Randomized trial of endobronchial ultrasound-guided transbronchial needle aspiration under general anesthesia versus moderate sedation. Am J Respir Crit Care Med. 2015;191(7):796–803. doi: 10.1164/rccm.201409-1615OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Yarmus L., Van der Kloot T., Lechtzin N., Napier M., Dressel D., Feller-Kopman D. A randomized prospective trial of the utility of rapid on-site evaluation of transbronchial needle aspirate specimens. J Bronchology Interv Pulmonol. 2011;18(2):121–127. doi: 10.1097/LBR.0b013e31821707ee. [DOI] [PubMed] [Google Scholar]
- 17.van der Heijden E.H., Casal R.F., Trisolini R. Guideline for the acquisition and preparation of conventional and endobronchial ultrasound-guided transbronchial needle aspiration specimens for the diagnosis and molecular testing of patients with known or suspected lung cancer. Respiration. 2014;88(6):500–517. doi: 10.1159/000368857. [DOI] [PubMed] [Google Scholar]
- 18.Wahidi M.M., Herth F., Yasufuku K. Technical aspects of endobronchial ultrasound-guided transbronchial needle aspiration: CHEST guideline and expert panel report. Chest. 2016;149(3):816–835. doi: 10.1378/chest.15-1216. [DOI] [PubMed] [Google Scholar]
- 19.Ong A.S.Q., Tan A.H., Anantham D. Impact of simulation training on performance and outcomes of endobronchial ultrasound-guided transbronchial needle aspiration performed by trainees in a tertiary academic hospital. J Thorac Dis. 2018;10(9):5621–5635. doi: 10.21037/jtd.2018.08.76. [DOI] [PMC free article] [PubMed] [Google Scholar]