Several techniques are available to manage central airway obstruction (CAO).1–3 These include airway dilation, tumor ablation or resection, and the insertion of endobronchial stents. Studies have shown the benefits of preintervention planning using computed tomography (CT).4–7 Imaging allows assessment of thoracic anatomy, identification of the sites of airway pathology, and quantification of the stenosis dimensions with the ultimate goal of selection of an appropriate therapeutic strategy, including correctly fitting prostheses (stents) when required.8
Several factors can limit the usefulness of preprocedural CT scanning in routine clinical practice. First, CT image quality may not be ideal. For example, scanning is not always done using state-of-the-art CT scanners with the ability of multiplanar reformatting. Even using modern machines and software algorithms, image quality may be degraded by patient movement or tachypnea, common in this population. In addition, scanning protocols vary, and are not all optimized for examining the central airways. Second, the scans may be done weeks or even months before the interventional procedure and rescanning may be impractical (eg, in urgent cases) or undesirable in situations in which radiation is of concern (eg, in children and nonmalignant CAO). During prolonged intervals between CT and bronchoscopy, tumors may grow and benign strictures can progress. Finally, the capacity of CT to accurately quantify functional disorders, such as tracheobronchomalacia, is limited to cine or paired inspiratory/expiratory scanning protocols and is missed with standard scans.
Regardless of image quality, bronchoscopists are required to estimate relevant airway dimensions during procedures to determine appropriate therapeutic interventions. At present, in vivo estimates of airway size are subjective and of variable accuracy. Interventional bronchoscopy is potentially hazardous and intraprocedural knowledge of airway dimensions could result in fewer adverse events from poorly fitting stents or overzealous airway dilation.9–13
Anatomical optical coherence tomography (aOCT), a validated light-based imaging technique, is suitable for such measurements.14 The endoscopic probe is passed through the working channel of a standard bronchoscope, and generates cross-sectional and 3-dimensional (3D) airway images from which dimensions can be determined with an unprecedented degree of precision. This case series examines the feasibility of incorporating aOCT airway imaging into a busy interventional pulmonology unit during bronchoscopic procedures for managing CAO. The utility of real-time stenosis measurements is also examined in light of the often variable quality of thoracic imaging available at the time of such procedures.
PATIENTS AND METHODS
The study was designed as a prospective case series.
Patients were recruited from the Interventional Pulmonology Unit of a university teaching hospital, having been referred for the management of symptoms related to CAO. Individuals in whom an interventional procedure was thought likely to alleviate symptoms were invited to participate in the study. The study was approved by the institutional Human Research Ethics Committee and informed written consent was obtained.
Our unit takes referrals from several local and regional hospitals. Accordingly, imaging was done on different scanners and provided in various formats including CD, hard copy, and, if performed on-site, the Picture Archiving and Communication System. The bronchoscopists (J.P.W. and M.J.P.), along with a thoracic radiologist (R.E.S.), measured the stenosed airway dimensions from the most recent scan. Stenosis length was determined from the point on the airway at which luminal narrowing commenced to the point at which the caliber normalized. Airway caliber immediately above and below the stenosis was measured. In the trachea, measurements were made in two planes (anteroposterior and transverse). These measurements guided therapeutic planning, such as choice of stent size or the anticipated degree to which an airway could be dilated.
Bronchoscopy was performed under general anesthesia while the patient breathed spontaneously through a laryngeal mask. Initial assessment was conducted using a flexible videobronchoscope. If required for stenting or for tumor ablation, a rigid bronchoscope was inserted. Techniques available for managing CAO included airway dilation using an inflated balloon or sequential bougie dilators, tumor ablation using Nd:YAG laser, argon plasma coagulation or snare diathermy, and airway stenting using Dumon (Novatech SA, La Ciotat, France), covered Ultraflex (Boston Scientific Corporation, Natick, MA), or covered Alveolus (Alveolus, Charlotte, NC) stents.
Anatomical Optical Coherence Tomography
OCT is a light-based imaging technique that measures the optical path length and amount of light backscattered from the tissues.15 Respiratory applications of OCT have been used to generate subsurface microscopic cross-sectional airway images.16,17 These “optical biopsies” typically have an axial scanning range of only 2 to 3 mm, as this is the depth to which the near-infrared light can usefully penetrate tissues and generate an OCT image.16,17 An adaptation of OCT, called aOCT, has increased the axial scanning range to allow hollow organs such as airways to be imaged in cross section. This aOCT system has been described in detail elsewhere.14,18–21 Briefly, it comprises a 1.3-mm-diameter fiberoptic probe housed within a clear plastic catheter (outer diameter 2.2 mm) as shown in Figure 1. The probe-catheter assembly is passed through the biopsy channel of a standard bronchoscope into an airway of interest (Fig. 2). On probe rotation (average frequency 2.5 Hz), broadband light (central wavelength 1310 nm) is projected onto the airway wall to acquire a 2-dimensional airway cross section that is displayed in real time on a monitor. The scanning probe can also be mechanically retracted across a region of interest (eg, an airway stenosis) at 0.4 mm/s and the data are reconstructed to generate 3D views of the lesion. This is referred to as a “pullback” scan. Immediately after the pullback scan, the aOCT data are analyzed with custom-designed software (Fig. 1) to provide measurements of relevant airway dimensions, and 3D visualizations are generated using a software viewer (VolView, Kitware, NY) (Fig. 3).
In each patient, a pullback scan was taken across the stenosis before and, for comparison, directly after the interventional procedure. In situations in which an endobronchial tumor required ablation, aOCT was used to measure the stenosis length and to assess the patency of the airway beyond the tumor. In situations in which the stenosis was very tight, or there was near-complete airway obstruction, the aOCT probe (Fig. 1) was carefully passed beyond the obstruction, taking great care not to provoke bleeding or airway perforation, in much the same way as a cytology brush may be passed beyond a tumor to obtain diagnostic tissue. Important airway dimensions were measured in the same way as for CT scanning and were based on the axial, coronal, and sagittal aOCT reconstructions (Figs. 3B–D).
Spirometry was performed before and after the procedure. In situations in which variable extrathoracic airway obstruction was present, this included measurement of forced inspiratory flow (FIF50%).
Performance Status and Dyspnea Scores
Performance status before and several weeks after the procedure was assessed using the Eastern Cooperative Oncology Group questionnaire.22 Subjective breathlessness was assessed using the MRC dyspnea score.23 Detailed case histories of 2 of these patients (patients 5 and 6) have been published earlier.21
Spirometry obtained before and after the procedures were compared using paired t tests. Performance status and dyspnea scores were compared using the Mann-Whitney rank-sum test. Data are presented as mean±standard deviation. Measurements of stenosis length and airway caliber made from CT and aOCT were compared using logistic regression and Bland-Altman analysis, a statistical function used to compare concordance and bias between the 2 measuring techniques.24 In a Bland-Altman plot, the y axis=CT−& aOCT and the x axis=(CT+aOCT)/2. The limits of agreement are specified as the values corresponding to 2 standard deviations of the difference. Differences not meeting this criterion are explained.
Fifteen aOCT studies were carried out in 14 patients (6 male) aged 59.6±16.0 years (patient 2 had 2 scans taken) (Table 1). Indications for bronchoscopy were malignant airway obstruction (n=6); benign laryngotracheal stenosis (n=4); tracheomalacia (TM) (n=2); radiation fibrosis (n=1); and endobronchial carcinoid (n=1). Interventional procedures included stent insertion (n=6); airway dilation (n=5); and tumor ablation (n=6). In 2 patients no intervention was performed; both had significant TM for which the risks of stenting outweighed the anticipated benefits.
Prebronchoscopy and postbronchoscopy spirometry was available and of satisfactory quality in 10 of the 12 patients who underwent interventional procedures. Forced expiratory flow in 1 second (FEV1) or FIF50% improved in 10 out of 10 of these patients. For the group, FEV1% predicted increased from 67±26% to 78±19% (P=0.04) and in the 4 patients with extrathoracic obstruction, FIF50% improved from 2.3±1.3 to 3.1±1.1 L/sec (P=0.09).
Dyspnea and Quality-of-Life Scores
Eastern Cooperative Oncology Group performance status improved in 10 of 12 patients after intervention: mean preprocedure score was 2.1±0.9 and improved to 1.1±0.9 (P=0.01). Dyspnea score improved in 9 of 12 patients after intervention: mean preprocedure score was 3.8±0.9 and improved to 2.9±0.8 (P=0.03).
CT and aOCT Imaging
The median time from CT scan to bronchoscopy was 3 weeks (range: 1 d to 26 mo). CT scans were of variable quality. Not all scans obtained outside our institution provided sufficient resolution or slice thickness to facilitate high-quality multiplanar reformatting. In these cases, the stenosis dimensions were based on measurements from axial slices. aOCT-based airway dimensions could be measured from 14 of the 15 pullback scans (Figs. 4, 5). Measurements could not be obtained around a tight bronchus intermedius stenosis in patient 2 because of excessive airway secretions, which reduced the aOCT contrast. CAO dimensions are compared in Table 2. CT and aOCT-based stenosis dimensions correlated closely (r 2=0.87, P<0.001) and Bland-Altman analysis showed a close agreement between the techniques, with mean difference 0.4±8.6 (2SD) mm (Fig. 6).
This study shows that an optical imaging technique, aOCT, can be incorporated into the interventional bronchoscopy suite to fulfill the currently unmet clinical need of providing real-time measurements of stenosis dimensions for bronchoscopic interventions, such as airway dilation and stenting. The clinical utility of such measurements is the provision of timely anatomical data on which treatment decisions can be based. In the majority of patients in this study (13/14), the data provided by aOCT aided the management strategy either by updating the stenosis dimensions when aggressive tumor growth had occurred between the CT and bronchoscopy, by confirming preexisting CT-based airway dimensions (on which stent selection was initially based), or by documenting the degree of dynamic airway collapse against which longitudinal disease progression could be assessed without the need for repeated exposure to ionizing radiation.
Bland-Altman analysis (Fig. 5) confirmed that CAO measurements based on aOCT are very similar to CT-based measurements. Several variables were uncontrolled between the measurements, including the time interval between CT and bronchoscopy, lung volume differences (awake CT versus anesthetized bronchoscopy), precise airway location, and phase of respiration. However, our results are consistent with those of an earlier study in which these variables were controlled and which confirmed that human in vivo CT and aOCT measurements compare closely.14
We have identified several areas in which aOCT may be of clinical benefit. The first relates to stent insertion for CAO. Typically, stent size is chosen based on preprocedure imaging and confirmed during bronchoscopy. Unfortunately, bronchoscope-based measurements are neither standardized nor accurate. A particular limitation is the optical distortion resulting from the “wide-angle” lens at the tip of the bronchoscope. Although techniques have been developed to overcome this distortion, none have been widely adopted as they are cumbersome and lack real-time capacity.8 The aOCT measurements in this study led to longer stents being used in 2 patients: patient 4 (because of underestimation of stenosis length using a visual bronchoscope-based technique) and patient 11 (because of rapid tumor growth in the 4 weeks after the CT scan). Optimizing stent length is important as stents that are too short for their intended function fail to maintain lumen patency, resulting in immediate or delayed airway occlusion. The second potential application of aOCT is in assisting airway dilation procedures. Tracheal caliber above and below the stenosis indicates normal airway caliber and knowledge of these measurements may help prevent overdilating the trachea and avoid attendant complications.25 A third application of aOCT is in the quantification and monitoring of the dynamic airway changes associated with TM. Patient 14 had an unsuspected moderately severe segment of TM not detected by CT but clearly seen at bronchoscopy and quantified by aOCT. As observation rather than intervention was considered the most appropriate strategy, aOCT would be an ideal tool to monitor subsequent progress of the TM without ionizing radiation. Finally, as shown in patient 5, the narrow aOCT probe can pass through tight stenoses to assess the patency of distal airways, an important consideration in determining treatment strategy and the likelihood of successful tumor ablation.26
We noted that despite the availability of state-of-the-art CT scanning facilities in our unit, several patients were transferred from other centers, but had preexisting scans already performed that were of insufficient quality for multiplanar reformatting. This limited the reliability of some CT-based measurements. The capacity to make these measurements during the bronchoscopy and avoid repeating the CT scan would, therefore, save time and resources and avoid further radiation exposure. Although on the whole, aOCT provided stenosis length measurements that were similar to those of CT (Fig. 6), there were exceptions (patients 5, 11, and 13). Each of these patients had aggressive malignant CAO and the CT scans were done between 3 to 8 weeks before the bronchoscopy, over which time significant tumor progression had occurred.
Limitations of Study Design and Use of aOCT
Excessive airway secretions and blood extinguish the aOCT near-infrared light, resulting in diminished airway wall definition (patient 2). Saline flushes and suctioning of secretions before pullback scans, if required, appeared to overcome this problem.
aOCT images were also affected by respiratory motion as pullback scans occurred over several breath cycles. This is indicated by the horizontal lines that appear on the 3D reconstructions indicating normal tracheal caliber during inspiration (Fig. 4) or accentuated caliber variations in TM (Fig. 6) and may be reduced by using respiratory gating techniques.27 Cough will also affect the aOCT images and an adequate depth of sedation or general anesthesia is therefore required to limit cough and optimize image quality.
Despite attempting to ensure a fair comparison of CT and aOCT measurements, 3 unavoidable factors may have influenced our results. First, it should be noted that the bronchoscopists were not blinded to the CT-based measurements as this was not considered ethically appropriate in the setting of exploration of a new technique. Second, defining the precise length of stenoses can be subjective, particularly when the narrowing is irregular, although this holds true for both CT and aOCT images. Third, suboptimal CT image quality at times limited the accuracy of airway measurements. Although a limitation for technique comparison, this often reflects clinical practice. Despite these limitations, the small mean difference between the techniques (0.4 mm) is considered clinically insignificant.
The addition of aOCT prolonged each procedure by approximately 10 to 15 minutes. Although undesirable for already lengthy procedures, further development of the aOCT system should allow faster pullback speeds at multiple airway sites to limit scanning time. Furthermore, although the postintervention pullback scans provided helpful images for illustrative purposes, these added little to patient management and could be omitted to further reduce scanning time.
Although this study shows that interventional procedures aided by aOCT result in overall outcomes consistent with published literature26,28 and do not increase the risk of the procedure, there is at present insufficient evidence to claim that interventions would be improved by the widespread uptake of the technique. Rather, this study shows the feasibility of incorporating aOCT into the bronchoscopy suite and paves the way for a study comparing functional and bronchoscopic outcomes in a larger cohort of patients randomized to aOCT-guided or non-aOCT-guided interventions.
We have described a novel imaging technique and shown its feasibility in the setting of interventional procedures to manage CAO. This technique provides accurate measurements of stenosis dimensions, is used through standard bronchoscopes, and complements CT imaging in planning treatment strategies and sizing airways before interventions. The study also highlights the need to perform intraprocedure measurements of airway dimensions when assessing airway stenoses, a task for which aOCT is an ideal tool, especially in a busy referral center at which off-site CT imaging data may at times be suboptimal.
The authors thank Drs Craig Smith and Markus Schmidt from the Department of Anaesthesia for their anesthetic expertise and Dr Vincent Low, Mr Neil Hicks, and Mr Peter Muir from the Department of Radiology, Sir Charles Gairdner Hospital, for assistance with CT imaging. The authors also thank the bronchoscopy nursing staff Sr Sue Morey, Sr Siobhan Dormer, and Mr John Crofts for their support and patience over the study duration.
1. Ernst A, Feller-Kopman D, Becker HD, et al. Central airway obstruction. Am J Respir Crit Care Med. 2004;169:1278–1297.
2. Wahidi MM, Herth FJF, Ernst A. State of the art-interventional pulmonology. Chest. 2007;131:261–274.
3. Williamson JP, Phillips MJ, Hillman DR, et al. Managing obstruction of the central airways. Intern Med J. 2010;40:399–410.
4. Boiselle PM, Lee KS, Ernst A. Multidetector CT of the central airways. J Thorac Imaging. 2005;20:186–195.
5. Heyer CM, Nuesslein TG, Jung D, et al. Tracheobronchial anomalies and stenoses: detection with low-dose multidetector CT with virtual tracheobronchoscopy-comparison with flexible tracheobronchoscopy. Radiology. 2007;242:542–549.
6. Rooney CP, Ferguson JS, Barnhart W, et al. Use of 3-dimensional computed tomography reconstruction studies in the preoperative assessment of patients undergoing balloon dilatation for tracheobronchial stenosis. Respiration. 2005;72:579–586.
7. Sun M, Ernst A, Boiselle PM. MDCT of the central airways: comparison with bronchoscopy in the evaluation of complications of endotracheal and tracheostomy tubes. J Thorac Imaging. 2007;22:136–142.
8. Williamson JP, James AL, Phillips MJ, et al. Quantifying tracheobronchial tree dimensions: methods, limitations and emerging techniques. Eur Respir J. 2009;34:1–14.
9. Alazemi S, Chatterji S, Ernst A, et al. Mediastinal migration of self-expanding bronchial stents in the management of malignant bronchoesophageal fistula. Chest. 2009;135:1353–1355.
10. Dooms C, De Keukeleire T, Janssens A, et al. Performance of fully covered self-expanding metallic stents in benign airway strictures. Respiration. 2009;77:420–426.
11. Murgu SD, Colt HG. Complications of silicone stent insertion in patients with expiratory central airway collapse. Ann Thorac Surg. 2007;84:1870–1877.
12. Park HY, Kim H, Koh WJ, et al. Natural stent in the management of post-intubation tracheal stenosis. Respirology. 2009;14:583–588.
13. Wadsworth SJ, Juniper MC, Benson MK, et al. Fatal complication of an expandable metallic bronchial stent. Br J Radiol. 1999;72:706–708.
14. Williamson JP, Armstrong JJ, McLaughlin RA, et al. Measuring airway dimensions during bronchoscopy using anatomical optical coherence tomography
. Eur Respir J. 2010;35:34–41.
15. Huang D, Swanson EA, Lin CP, et al. Optical coherence tomography
. Science. 1991;254:1178–1181.
16. Coxson HO, Quiney B, Sin DD, et al. Airway wall thickness assessed using computed tomography and optical coherence tomography
. Am J Respir Crit Care Med. 2008;177:1201–1206.
17. Whiteman SC, Yang Y, van Pittius DG, et al. Optical coherence tomography
: real-time imaging of bronchial airways microstructure and detection of inflammatory/neoplastic morphologic changes. Clin Cancer Res. 2006;12:813–818.
18. Armstrong JJ, Leigh MS, Walton ID, et al. In vivo size and shape measurement of the human upper airway using endoscopic long-range optical coherence tomography
. Opt Express. 2003;11:1817–1826.
19. Leigh MS, Armstrong JJ, Paduch A, et al. Anatomical optical coherence tomography
for long-term, portable, quantitative endoscopy. IEEE Trans Biomed Eng. 2008;55:1438–1446.
20. McLaughlin RA, Williamson JP, Phillips MJ, et al. Applying anatomical optical coherence tomography
to quantitative 3D imaging of the lower airway. Opt Express. 2008;16:17521–17529.
21. Williamson JP, McLaughlin RA, Phillips MJ, et al. Using optical coherence tomography
to improve diagnostic and therapeutic bronchoscopy. Chest. 2009;136:272–276.
22. Oken MM, Creech RH, Tormey DC, et al. Toxicity and response criteria of the Eastern-Cooperative-Oncology-Group. Am J Clin Oncol Cancer Clin Trials. 1982;5:649–655.
23. Fletcher CM, Elmes PC, Fairbairn AS, et al. The significance of respiratory symptoms and the diagnosis of chronic bronchitis in a working population. Br Med J. 1959;2:257–258.
24. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307–310.
25. Kim JH, Shin JH, Song HY, et al. Tracheobronchial laceration after balloon dilation for benign strictures-incidence and clinical significance. Chest. 2007;131:1114–1117.
26. Cavaliere S, Venuta F, Foccoli P, et al. Endoscopic treatment of malignant airway obstructions in 2008 patients. Chest. 1996;110:1536–1542.
27. McLaughlin RA, Armstrong JJ, Becker S, et al. Respiratory gating of anatomical optical coherence tomography
images of the human airway. Opt Express. 2009;17:6568–6577.
28. Breitenbuecher A, Chhajed PN, Brutsche MH, et al. Long-term follow-up and survival after Ultraflex stent insertion in the management of complex malignant airway stenoses
. Respiration. 2008;75:443–449.
Keywords:© 2010 Lippincott Williams & Wilkins, Inc.
airway stenoses; optical coherence tomography; interventional bronchology; lung cancer