Lung cancer is a leading cause of cancer death, and its early identification is imperative to guide treatment. Diagnostic evaluation should ideally be performed using the most accurate, least invasive, and least expensive approach. Currently, the method for diagnosing lung cancer depends upon a variety of factors, such as the size, location, and morphology of the suspected lesion; the presence of metastasis; and the clinical history and status of the patient. When the presence of malignancy is highly likely, the best option might be surgical biopsy and resection. Often, however, the probability of malignancy is uncertain, and moreover, some lung lesions are nonresectable. In such cases, additional diagnostic methods are mandatory. The diagnostic procedure of choice, when the target can be reached and an adequate tissue sample can be obtained, is flexible bronchoscopy (FB), which is a minimally invasive, and relatively low cost procedure.1 Lung masses, however, are often located in the airways beyond the vision of conventional FB.2,3
When a definitive diagnosis is not provided by bronchoscopy, the patient might require repeat of the procedure, or undergo more invasive, higher-risk procedures, such as computed tomography (CT)-guided percutaneous needle biopsy, or surgical biopsy and resection.4,5 Occasionally in cases of small apparently benign lesions, the management strategy is “watchful waiting” with repeat CT scans until a definitive diagnosis can be made.6
Conventional transbronchial needle aspiration (TBNA) of the lymph nodes for cancer staging also has limitations due to the lack of real-time visualization, which results in difficulty navigating to the target location. The physician reaches the “invisible” lymph node through the airway wall, on the basis of the knowledge of anatomy, and memorization of the patient's CT image.7 More recently, endobronchial ultrasound-driven biopsy has been successfully used to increase the diagnostic yield for peripheral lung lesions and lymph nodes.8–11
The superDimension electromagnetic, image-guided localization and navigation system [electromagnetic navigation bronchoscopy (ENB), superDimension, Ltd, Herzliya, Israel] is designed to facilitate reaching small suspicious peripheral lung lesions that are outside the reach of standard bronchoscopes. The electromagnetic navigation (EMN) system extends beyond the reach of the FB and provides an alternative option for a minimally invasive diagnosis of lung lesion. Details of the components and use of this system were previously published.12–15
In addition to this novel technology, we have used rapid on-site cytologic evaluation (ROSE) of transbronchial aspirates. This technique is safe and improves the diagnostic rate of standard bronchoscopy, because it overcomes the problems of nondiagnostic negative specimens, such as inadequate specimens, and those that contain only mucus, respiratory epithelium, or blood.16–20 However, the value of its routine use for diagnosis of peripheral lung lesions has not been clearly established. In our institution, ROSE is used as a routine practice. The purpose of this single-center, retrospective study was to evaluate the diagnostic yield of bronchoscopy, guided by the EMN system combined with the ROSE technique. The main emphasis of this study was to determine the proportion of patients that were able to receive a malignant or plausible nonmalignant diagnosis the day of the procedure.
EMN is routinely used at our institution to perform nonemergency biopsies of suspicious peripheral lung lesions and TBNA procedures of enlarged mediastinal lymph nodes that are considered difficult to reach using a standard bronchoscopic approach. All consecutive patients who had procedures with the EMN between June 2005 and July 2006 were included in the study. All patients were under the care of the first author who performed all EMN procedures. The data collection for this retrospective study was conducted by the authors in compliance with the health information privacy regulations in the Health Care Portability and Accountability Act of 1996. Permission was obtained from the Institutional Review Board to carry out the data collection.
The diagnostic test reports and patient medical records were retrospectively reviewed by the authors, and detailed information was collected regarding the EMN performance, safety, and diagnostic yield. The majority of patients were followed up for confirmation of final diagnosis and/or treatment. When clinically indicated, additional follow-up procedures, such as surgery, mediastinoscopy, or CT-guided, fine-needle aspiration were performed. Patients who were lost to follow-up, or referred to other hospitals for clinical follow-up, were classified as having an inconclusive diagnosis.
The primary objective of the study was to measure diagnostic yield per patient on the day of the procedure. Secondary objectives included the assessment of diagnostic yield on the basis of any additional diagnostic procedures (per sampled lung target and per patient) and overall performance and safety of EMN-guided bronchoscopy in a community based practice.
Cases that led to the diagnosis of malignant or nonmalignant disease were considered as diagnostic on the day of procedure. Cases that did not lead to the diagnosis of disease were considered as nondiagnostic on the day of procedure. These nondiagnostic cases were further followed-up by additional diagnostic methods, and thus, became either diagnostic or nondiagnostic, on the basis of the follow-up information. Cases, for which no follow-up information was available, were considered as inconclusive. Sensitivity analysis was performed to estimate the diagnostic yield under 4 assumptions regarding the inconclusive cases. First, all inconclusive cases were treated as nondiagnostic (false negative, worst-case scenario). Second, all inconclusive cases were treated as diagnostic (true negative, best-case scenario). Third, the estimate of the percent of diagnostic and nondiagnostic cases from the observed data was applied to the inconclusive cases. Finally, all inconclusive cases were excluded from the analysis.
Overall performance of the EMN system was evaluated in terms of the ability to successfully navigate to the lung target with EMN guidance; the distance from the tip of the steerable, sensor-probe to the center of the peripheral lung lesions, and the accuracy of the registration process, as measured by the average fiducial target registration error (AFTRE). Safety of the EMN system was assessed by documenting all complications reported during the procedures, both related and unrelated to the system.
All subjects had CT examinations configured with 2-mm-thick slices at 1-mm intervals in Digital Imaging and Communications in Medicine (DICOM) format. CT images were uploaded into the superDimension system, which uses the images to reconstruct a 3-dimensional virtual bronchoscopy of the patient's lungs. EMN was then used to locate, register, and navigate to lung targets in accordance with the manufacturer's instructions, as previously described12–15 (Fig. 1). All procedures were performed using a conscious sedation protocol with intravenous Versed (midazolam·HCl; 2 mg), fentanyl (50 mcg), and titrated propofol dosing. Local anesthesia consisted of topical lidocaine in the nasopharynx and oropharynx as required throughout the course of the procedure. Fluoroscopic guidance was used after the completion of navigation to the lung target to verify the accuracy of navigation, and optimize the results of the biopsy and TBNA procedures.
Tissue sampling via lung tissue biopsy or TBNA was performed using standard endoscopic tools. The tools included a standard bronchoscope (BF1T 160, Olympus America Inc, Center Valley, PA), biopsy forceps (Gastropediatric UPN M00515780, Boston Scientific, Global Park, Heredia, Costa Rica), and needles (WANG Transbronchial Cytology Needle, 21 g, Ref SW-521, WANG Transbronchial Histology Needles, 21 g and 19 g, Ref MW-319 and WANG Transbronchial Histology Needle, 19 g, Ref MWF-319, BARD Endoscopic Technologies, Billerica, MA). On an average, approximately 3 to 4 samples were obtained by forceps and needle from lung lesions, and approximately 5 to 6 samples by needle from lymph nodes. ROSE was used almost in all cases, which is standard practice at our institution.
Descriptive statistics for all continuous variables were summarized as means, SDs, and ranges. Percentages, counts, and 95% confidence intervals were reported for discrete variables. Analyses were performed using commercially available statistical analysis software (SPSS, version 12.1; SPSS Chicago, IL)21 by the Statistical Laboratory, Department of Statistics and Operation Research, School of Mathematical Sciences, Tel Aviv University, Tel Aviv, Israel. The following parameters were assessed: patient demographics, lung target size and location, distance from the tip of the steerable sensor-probe to the center of the peripheral lung lesion, registration accuracy, navigation success, diagnostic yield per lung target and per patient, and procedural complications. For statistical tests of association (independent t test and χ2), a P value of ≤0.05 was considered significant.
The study included 248 patients (122 men, 126 women; ages 30 to 91 y; mean age 63.1±12.9 y). From these 248 patients, 279 peripheral lung lesions, and 71 lymph nodes were targeted for diagnosis. Only peripheral lung lesions were targeted in 190 patients, only lymph nodes in 26 patients, and both were targeted in 32 patients. Available follow-up information was recorded for 6 to 18 months (mean 6±5 mo) after the procedure. Clinical follow-up was recorded for 181 patients; 67 patients did not have clinical follow-up available.
Lung Target Characteristics
The size of the targeted peripheral lung lesions ranged from 1.1 to 7.3 cm, with a mean of 2.1±1.4 cm. The size of the targeted lymph nodes ranged from 1.5 to 3.6 cm, with a mean of 1.8±0.9 cm. The distributions of location and size of the peripheral lung lesions and lymph nodes, along with related navigation success are presented in Table 1. The majority (51%) of targeted peripheral lung lesions were present in the upper lung lobes, which is considered the most difficult region to reach by bronchoscopy.22
Performance and Safety
Navigation success was defined as the ability of the EMN system to reach the lung target. At least one lung target was reached successfully in all 248 patients (100%). A total of 266/279 (95.3%) peripheral lung lesions and 67/71 (94.3%) lymph nodes were successfully reached, and tissue samples for biopsy were obtained. The remaining 13 peripheral lung lesions and 4 lymph nodes were not reached for the following reasons:
- Four lymph nodes and 8 peripheral lung lesions-although created as targets during the planning phase, the investigator decided not to attempt to reach them because satisfactory results were obtained with other targets.
- Five peripheral lung lesions were not reached despite several attempts due to limited access to the lesion.
Navigation success was similar for all lobar and mediastinal target locations (P>0.05) (Table 1).
The accuracy of the registration process was assessed with the AFTRE score, which is automatically calculated by the system software. It was defined as the radius of the difference of the location of the tip of the steerable sensor-probe in the actual patient, compared with where it was expected to be on the basis of the information of the virtual patient (ie, CT). The AFTRE score ranged from 0.2 to 1.1 cm with a mean of 0.5±0.02 cm. AFTRE scores of less than 5 mm are considered acceptable.12
The distance between the tip of the steerable sensor-probe and the center of the targeted peripheral lung lesions ranged from 0.12 to 3.0 cm, with a mean of 0.8±0.5 cm. This distance, combined with the small size of the targeted peripheral lung lesions (mean 2.1±1.4 cm), suggests that, in most of the procedures, the biopsy sample was obtained directly from the lesion itself. It was not possible to determine the comparable distance measurement for lymph nodes because the sensor-probe is placed on the trachea adjacent to the target lymph node.
There were 8 complications: 3 cases of moderate bleeding (none requiring transfusion), 3 cases of pneumothorax that resolved without treatment, 1 case of hematoma that also resolved without treatment, and 1 case of pneumonia/chronic obstructive pulmonary disease exacerbation that was treated on an outpatient basis with oral antibiotics. None of these events were directly related to the use of the EMN system. Expected, temporary, nonserious, procedure-related complications of biopsy and TBNA procedures, such as cough, sore throat, or chest pain, were not recorded for this study.
Tables 2A and B present the diagnostic yield results. According to our results, 161/248 (65%) of patients left the bronchoscopic procedure with a definitive diagnosis. On the basis of the additional clinical follow-up and diagnostic procedures via CT, another 12 patients (5%) were confirmed as negative, over time. Thus, 70% of all patients had a definitive diagnosis the day of the procedure that remained true over time. This estimate of 70% represents the worst-case scenario, when all inconclusive cases are considered as nondiagnostic. When all inconclusive cases are treated as diagnostic (best-case scenario) the estimate is 97%. Thus, the diagnostic yield ranges between 70% and 97% based upon the assumptions made regarding the outcome of the cases that had an inconclusive diagnosis on the day of the procedure. When the estimate of the percent of diagnostic and nondiagnostic cases from the observed data was applied to the inconclusive cases, the diagnostic yield was 86%. Table 2A displays the diagnoses obtained with the aid of EMN technology on the day of procedure. Table 2B displays number and percentage of diagnostic and nondiagnostic cases on the day of procedure, and during follow-up with the added information from the results of additional diagnostic procedures. Diagnostic yield was similar for all lung target locations (P>0.05).
The results of this retrospective study of 248 patients are consistent with previous reports of the relatively high diagnostic yields obtained using the EMN biopsy system compared with conventional bronchoscopy.12–14 The diagnostic yields obtained in this study (65% the day of the procedure, with at least 70% proven over time) represent an improvement over diagnostic techniques that either take more time for obtaining a conclusive result or are more invasive. Sensitivity analyses to adjust for loss-to-follow-up suggest the diagnostic yield may be as high as 86%. The combined approach of using EMN along with the ROSE technique contributed to the increased yield of bronchoscopy observed in this study.
The high incidence of granuloma and benign inflammation in this cohort is notable. This observation was expected and is partially explained by the geographic location of the study population (Indiana), in which there is a high incidence of histoplasmosis.23
Gildea et al12 reported a diagnostic yield of 74% for peripheral lesions with ENB in the academic hospital setting. The yield of 74% obtained by Gildea et al is comparable to the 70% yield over time obtained in our community setting, despite differences in patient population and diagnostic techniques between the 2 institutions. The patient population in our study had a lower prevalence of malignant disease (37% vs. 77%). In the present study, CT scan slices were 2-mm thick at 1-mm intervals in the DICOM format, versus 3-mm thick at 1.5-mm intervals in Gildea et al12; thinner slices might increase diagnostic yield.24 The biopsy technique used was different in the present study, as tissue samples were obtained with needles and forceps rather than brushings. Finally, ROSE was used for performing all biopsies in the present study, which might also account for the higher yield. The increased diagnostic yield with ROSE has been previously reported by several authors (up to 90.3% for peripheral lesions).16–18
There were no unexpected or added risks associated with the use of EMN. Procedural complications were very low (3.2%), and compare favorably to published reports of complications of transbronchoscopic biopsy, TBNA, and EMN.12–14,25,26
This study represents the largest series of patients (n=248) evaluated for diagnostic yield with EMN. It provides the best estimate available of diagnostic yield with EMN in a community setting with a relatively high prevalence of benign disease in patients.
As a retrospective study, there are a number of potential limitations. Selection bias might limit the generalizability of the results. As already noted, differences in the patient population might partially explain variations in the diagnostic yield across studies. Because the length of clinical follow-up was 6 to 18 months for patients with negative diagnoses, it is possible that with further follow-up, malignant disease might be found in some of these patients. Finally, there were 67 patients with no clinical follow-up available (27%) and their diagnostic status remained inconclusive.
In conclusion, the results of this study confirm that EMN is a safe, effective, and noninvasive alternative for the diagnosis of peripheral lesion. Use of EMN might reduce the necessity for more invasive, higher risk diagnostic procedures, or unnecessary surgery. Its use in combination with ROSE technique may significantly increase the diagnostic yield of routine bronchoscopy. Prospective studies with longer clinical follow-up, and studies of the impact of EMN and ROSE use on clinical decision making, patient management, and patient outcomes are needed to further elaborate the value of our data.
1. Alberts WM, Colice GL. Diagnosis and management of lung cancer
: ACCP evidence-based guidelines. Chest. 2003;123(suppl 1):1S–337S.
2. Rivera MP, Detterbeck F, Mehta AC. American College of Chest Physicians. Diagnosis of lung cancer
: the guidelines. Chest. 2003;123(suppl 1):129S–136S.
3. Funakoshi Y, Sawabata N, Takeda S, et al. Bronchoscopically undiagnosed small peripheral lung tumors. Interact Cardiovasc Thorac Surg. 2003;2:517–520.
4. Stanley JH, Fish GD, Andriole JG, et al. Lung lesions: cytologic diagnosis by fine-needle biopsy. Radiology. 1987;162:389–391.
5. Venn G, Goldstraw P. Open lung biopsy. Br J Hosp Med. 1988;39:272–276.
6. Yankelvitz DF, Henschke CI. Does 2-year stability imply that pulmonary nodules are benign? Am J Roentgenol. 1997;168:325–328.
7. Kramer H, Groen HJ. Current concepts in the mediastinal lymph node staging on non-small cell lung cancer
. Ann Surg. 2003;238:180–188.
8. Sachs S, Bilfinger TV. The impact of positron emission tomography on clinical decision making in a university-based multidisciplinary lung cancer
practice. Chest. 2005;128:698–703.
9. Chhajed PN, Bernasconi M, Gambazzi F, et al. Combining bronchoscopy
and positron emission tomography for the diagnosis of small pulmonary nodule ≤3 cm. Chest. 2005;128:3558–3564.
10. Herth FJ, Lunn W, Eberhardt R, et al. Transbronchial versus transesophageal ultra-sound guided aspiration of enlarged mediastinal lymph nodes
. Am J Crit Care Med. 2005;171:1164–1167.
11. Paone G, Nicastri E, Lucantoni G, et al. Endobronchial ultrasound-driven biopsy in the diagnosis of peripheral lung lesions. Chest. 2005;128:3551–3557.
12. Gildea TR, Mazzone PJ, Karnak D, et al. Electromagnetic navigation diagnostic bronchoscopy
: a prospective study. Am J Respir Crit Care Med. 2006;174:982–989.
13. Schwarz Y, Greif Y, Becker H, et al. Real-time electromagnetic navigation bronchoscopy
to peripheral lung lesions using overlaid CT images: the first human study. Chest. 2006;129:988–994.
14. Becker H, Herth F, Ernst A, et al. Bronchoscopic biopsy of peripheral lung lesion under electromagnetic guidance: a pilot study. J Bronchol. 2003;12:9–13.
/Bronchus User Manual, DUM00021-01, Revision D.
16. Baram D, Garcia RB, Richman PS. Impact of rapid on-site cytologic evaluation during transbronchial needle aspiration. Chest. 2005;128:869–875.
17. Uchida J, Imamura F, Takenaka A, et al. Improved diagnostic efficacy by rapid cytology test in fluoroscopy-guided bronchoscopy
. J Thorac Oncol. 2006;1:314–318.
18. Davenport R. Rapid on-site evaluation of transbronchial aspirates. Chest. 1990;98:59–61.
19. Yung RC. Tissue diagnosis of suspected lung cancer
: selecting between bronchoscopy
, transthoracic needle aspiration, and resectional biopsy. Respir Care Clin North Am. 2003;9:51–76.
20. Diacon AH, Schuurmans MM, Theron J, et al. Utility of rapid on-site evaluation of transbronchial needle aspirates. Respiration. 2005;72:182–188.
21. SPSS for Windows, Release 12.1. 2003. Chicago: SPSS Inc.
22. Sakurada A, Takahashi N, Sato M, et al. Are difficulties during transbronchial lung biopsy/brushing through a fiberoptic bronchoscope based on bronchial anatomy? Surg Radiol Anat. 2005;27:94–99.
24. Iwano S, Makino N, Ikeda M, et al. Solitary pulmonary nodules: optimal slice thickness of high resolution CT in differentiating malignant from benign. Clin Imaging. 2004;28:322–328.
25. Lukomsky GI, Ovchinnikov AA, Bilal A. Complications of bronchoscopy
: comparison of rigid bronchoscopy
under general anesthesia and flexible fiberoptic bronchoscopy
under topical anesthesia. Chest. 1981;79:316–321.
26. Mazzone P, Jain P, Arroliga AC, et al. Bronchoscopy
and needle biopsy techniques for diagnosis and staging of lung cancer
. Lung Cancer
Keywords:© 2007 Lippincott Williams & Wilkins, Inc.
bronchoscopy; community hospital; diagnostic imaging; electromagnetic navigation bronchoscopy (ENB); lung cancer; mediastinal lymph nodes; rapid on-site cytologic evaluation (ROSE); solitary pulmonary nodule; superDimension