Lung

, 187:55

Electromagnetic Navigation Bronchoscopy in Combination with PET-CT and Rapid On-site Cytopathologic Examination for Diagnosis of Peripheral Lung Lesions

  • Bernd Lamprecht
  • Peter Porsch
  • Christian Pirich
  • Michael Studnicka
Article

DOI: 10.1007/s00408-008-9120-8

Cite this article as:
Lamprecht, B., Porsch, P., Pirich, C. et al. Lung (2009) 187: 55. doi:10.1007/s00408-008-9120-8

Abstract

Background The combination of electromagnetic navigation bronchoscopy (ENB), PET-CT, and rapid on-site cytopathologic examination (ROSE) for the routine diagnostic work-up of peripheral lung lesions has not been evaluated previously. Objectives The aim of this study was to determine the accuracy and safety of ENB in combination with PET-CT and ROSE in subjects with endobronchially invisible peripheral lung lesions. Methods ENB was performed in 13 subjects with radiologically suspected lung cancer who were referred to our tertiary-care hospital between October 2005 and November 2006. ENB was performed using the superDimension/Bronchus System. FDG-PET-CT scans were part of the diagnostic workup. Bronchoscopy was done under general anesthesia and ROSE was available in this setting. The final diagnosis was based on the histopathologic results of specimens obtained either by ENB or, if ENB was not diagnostic, by surgery or CT-guided fine-needle aspiration (FNA). Results The mean diameter of peripheral lesions ranged from 1.4 to 5.3 cm (average = 3.0 ± 1.2 cm). In 76.9% of the patients, ENB resulted in obtaining a correct diagnosis, as defined by the definite histopathologic result. Sensitivity and specificity of ROSE was 84.6 and 100%, respectively. In malignant lesions the SUV ranged from 2.0 to 17.0 and was independent of lesion size. The positive predictive value of a positive PET-CT scan for a diagnosis of malignancy was 90%. No ENB-related adverse events were seen during and up to 24 h after bronchoscopy. Conclusion ENB in combination with PET-CT and ROSE is safe and effective in the diagnostic workup of peripheral lung lesions.

Keywords

Bronchoscopy Electromagnetic navigation bronchoscopy Peripheral lung lesion Transbronchial biopsy PET-CT 

Introduction

The management of pulmonary nodules gains attention as the incidence of these lesions increases with utilization of more sensitive diagnostic methods. In high-risk populations, screening for lung cancer using low-dose CT has shown an incidence of suspicious lung nodules of 2.2% [1].

Sensitivity of routine bronchoscopy for the diagnosis of peripheral lesions less than or greater than 2 cm is reported to range from 33 to 62%, respectively [2]. For lesions with a diameter of less than 2 cm located in the periphery of the lung, diagnostic yield decreases to 14% [3]. In this situation adjuvant bronchoscopic technologies such as fluoroscopic guidance, endobronchial ultrasound (EBUS), and electromagnetic navigation bronchoscopy (ENB) can help improve sensitivity. In peripheral lung lesions it has been shown that electromagnetic navigation bronchoscopy increases diagnostic yield 59–74% [4, 5, 6, 7, 8].

Inadequate specimen collection is another factor limiting the yield of bronchoscopy. In this context, rapid on-site examination of transbronchial biopsies has been reported to be very useful, accurate, and cost effective [9]. In non-small-cell lung cancer the maximum standardized uptake value (SUV) of pulmonary nodules in positron emission tomography has been shown to be a predictor of stage and tumor characteristics [10]. Therefore, the use of PET/CT in the diagnostic approach to peripheral lung lesions helps to plan diagnostic and therapeutic procedures.

In this study PET/CT was performed before the bronchoscopic procedure and ROSE was used to overcome the limitation of inadequate specimen collection. This is the first study on the diagnostic yield of ENB (using superDimension/Bronchus System) in a clinical setting with adjunct ROSE and PET-CT preceding bronchoscopy.

Methods

Clinical Setting

The data were recorded at the Department of Pulmonary Medicine, Paracelsus Medical University, Salzburg, Austria. This is a tertiary-care unit in which more than 500 bronchoscopies per year are performed. The majority of these procedures are performed in subjects referred for suspected lung cancer. In our clinical setting rapid on-site cytopathologic examination (ROSE) and PET-CT are part of the routine diagnostic workup.

Study Population

Thirteen consecutive patients who underwent ENB between October 2005 and November 2006 were included in the study. Patient records were retrospectively reviewed to determine the performance of ENB in combination with PET-CT and ROSE, diagnostic yield, and confirmed final diagnosis. All subjects were candidates for nonemergency bronchoscopy of a suspected peripheral lung lesion. All presented with lesions traditionally not reachable by routine bronchoscopy (located in the peripheral third of the chest and/or to small to be visible on chest radiograph and uniplanar fluoroscopy during bronchoscopy). All subjects provided informed consent before undergoing ENB.

FDG-PET-CT

Before ENB subjects underwent a chest CT scan configured with 3-mm-thick slices at 1.5-mm intervals. Twelve of 13 underwent integrated fluorodeoxyglucose-positron-emission-computed tomography (FDG-PET-CT). PET/CT imaging systems combine the functional sensitivity of PET with the anatomical detail of multislice CT. In lung cancer (18)F FDG-PET-CT is a valuable discriminator of disease load. For each patient the maximum standard uptake value (maxSUV) of the endobronchially invisible lesion was recorded. A SUV ≥ 2 was assumed to predict a malignant diagnosis. The preoperative CT data were used by the superDimension/Bronchus System software to generate a 3-dimensional CT roadmap.

Electromagnetic Navigation Bronchoscopy

All bronchoscopic procedures were performed by the same operator. The electromagnetic navigation system is an image-guided localization device that assists the endobronchial accessories (forceps, needle, and brush) in reaching the endobronchially invisible peripheral lung lesions. The electromagnetic navigation system consists of an electromagnetic location board, a steerable sensor probe (locatable guide), an extendable working channel, and computer software. During bronchoscopy the steerable sensor probe is advanced toward the target. Details of the equipment and configuration have already been described [11, 12]. All procedures were done using the rigid bronchoscopic intubation technique and under general anesthesia using an Olympus 1T160, 2.8-mm working channel, adult therapeutic bronchoscope. No additional guidance technique (e.g., fluoroscopic guidance) was used.

Rapid On-site Cytopathologic Evaluation (ROSE)

Bronchoscopy was done in combination with rapid on-site cytopathologic examination, which is routinely available at our institution. ROSE is facilitated by immediate smearing of the specimens onto slides, drying, and fixation. An experienced cytopathologist evaluated specimens sampled by variable biopsy techniques (brush, forceps, needle). The Papanicolaou (Pap) test grading system (I-V) was used to describe the findings of the cytopathologic examination. For confirmation of final diagnosis, a histopathologic examination was performed in all cases.

Statistical Analysis

The parametric t test was used to evaluate the impact of lesion size on the diagnostic yield of electromagnetic navigation bronchoscopy. All statistical analyses were done with SAS 8.2 (SAS Institute Inc., Cary, NC).

Results

Electromagnetic navigation bronchoscopy was performed in all 13 patients; the results are summarized in Table 1. The size of the peripheral lung lesions ranged from 1.4 to 5.3 cm (average = 3.0 ± 1.2 cm). The location of the peripheral lesions was as follows: two lesions were located at the left lower lobe (LLL), three at the left upper lobe (LUL), two at the right lower lobe (RLL), five at the right upper lobe (RUL), and one at the right middle lobe (RML).
Table 1

Characteristics of patients and peripheral lung lesions

Number of patients (lesions)

13 (13)

Age (years) (mean ± SD)

64.2 ± 10.7

Male/Female

10/3

Diameter of lesion (cm) (mean ± SD)

3.0 ± 1.2

Final diagnosis (%)

    Primary lung cancer

61.5%

    Metastatic lung cancer

7.7%

    Benign disease

30.8%

The duration of general anesthesia, including the time for the total bronchoscopic procedure, ranged from 20 to 85 min with a mean time of 60 min. In 76.9% (10/13) ENB procedures resulted in a correct diagnosis (see Table 2). No patient had more than five biopsies to establish the diagnosis (median was three biopsies).
Table 2

Location, size, rapid on-site cytopathologic evaluation results, and histopathologic biopsy results by electromagnetic navigation bronchoscopy

Case no.

Lesion location

Lesion size (cm)

FDG-PET-CT maxSUV

ROSE (specimen by ENB)

Histopathologic result (specimen by ENB)

1

LLL

1.4

2.6

PAP II

Normal tissue [A]a

2

RUL

1.6

0.0

PAP III

Hamartoma

3

RUL

5.3

17.0

PAP V

NSCLC (adenocarcinoma)

4

RML

3.8

No PET

PAP II

Lung parenchymal involvement by vasculitis

5

LLL

2.0

8.4

PAP II

NSCLC (adenocarcinoma)

6

RLL

3.3

2.0

PAP III

Normal tissue [B]a

7

RUL

3.2

5.0

PAP V

NSCLC (unspecified)

8

LUL

2.8

6.6

PAP V

NSCLC (adenocarcinoma)

9

RUL

5.2

7.7

PAP III

Mycobacterial infection (nontuberculosis)

10

LUL

2.0

8.0

PAP II

SCLC

11

RUL

3.0

9.4

PAP V

NSCLC (squamous cell Ca)

12

LUL

2.6

2.5

PAP II

Normal tissue [C]a

13

RLL

2.8

0.0

PAP II

Hamartoma

LLL = left lower lobe; LUL = left upper lobe; RLL = right lower lobe; RUL = right upper lobe; RML = right middle lobe

aFinal diagnosis by surgery/FNA: [A] metastasis of breast cancer, [B] squamous cell carcinoma, [C] bronchoalveolar cell carcinoma

Three of nine (33.3%) diagnoses established by ENB were false negative. These cases were finally diagnosed by surgery (n = 2) or by CT-guided fine-needle aspiration (n = 1). One was a metastasis of breast cancer, one was a squamous cell carcinoma, and one was a bronchoalveolar cell carcinoma. Diagnostic yield of ENB was not significantly different (p = 0.379) by lesion size.

The result of ROSE and the definite histopathologic result of the lesion biopsy were concordant in 11 of 13 cases (84.6%). In five of nine cases (55.6%) ROSE was false negative. The sensitivity and specificity of ROSE were 84.6 and 100%, respectively.

The standard uptake value (SUV) recorded during PET-CT was negative in the two benign pulmonary lesions (hamartoma). In malignant lesions the SUV ranged from 2.0 to 17.0 and was independent of lesion size. In a case of nontuberculous mycobacterial infection, the detected SUV was 7.7. Overall, PET-CT showed a sensitivity and specificity for the correct diagnosis of malignant lesions (primary lung cancer, metastatic lung cancer) of 100 and 66.7%, respectively. The positive and negative predictive values for a diagnosis of malignancy are shown in Table 3.
Table 3

Establishment of a correct diagnosis of malignancy in peripheral lung lesions by PET-CT, ENB, and ROSE

Method

Correct diagnosis of malignancy (%)

False-negative rate (%)

False-positive rate (%)

PPV (%)

NPV (%)

PET-CTa (n = 12)

91.7

0

33.3

90

100

ENB (n = 13)

76.9

33.3

0

100

57.1

ROSEb (n = 13)

84.6

55.6

0

100

66.7

PPV = positive predictive value; NPV = negative predictive value; False negative, False positive, PPV and NPV refer to a diagnosis of malignancy

aSUV ≥ 2.0

bFor lesions definitely reached by ENB

In the two cases of malignant lesions in which ROSE showed a false-negative result, PET-CT reported an elevated SUV (8.4 and 8.0) and histopathologic examination confirmed the diagnosis of malignancy. A flowchart of the diagnostic procedure is shown in Fig. 1.
Fig. 1

Flowchart of diagnostic workup

No pneumothorax and no other ENB-related adverse events were seen during and up to 24 h after the procedure.

Discussion

In this series of patients with endobronchially invisible pulmonary lesions, electromagnetic navigation bronchoscopy in combination with rapid on-site cytopathologic examination and prior PET-CT was shown to be safe and highly effective. In small peripheral lung lesions, a diagnostic yield of 77% is an excellent result. Recently published studies on ENB have shown similar results with diagnostic yields between 59% and 74% [5, 6, 7, 8]. The combination of flexible standard bronchoscopy with uniplanar fluoroscopy and PET scan has shown a diagnostic yield of 53% in lesions less than 3 cm [13].

Electromagnetic navigation has increased the sensitivity of bronchoscopy close to the diagnostic yield that is reported for transthoracic CT-guided (92%) [2] or surgical (nearly 100%) [14] biopsies. In contrast to these more invasive procedures which have several risks [15, 16], bronchoscopy has less risks [17] and is usually more comfortable. In the patients studied at our institution no adverse effects of ENB were seen. Previous studies reported an incidence of pneumothorax between none and 8% [5, 6, 7, 8]. This is remarkably lower than the pneumothorax rate reported for CT-guided coaxial-cutting needle biopsies (23%) [15]. Besides the greater risk of pneumothorax and bleeding, transthoracic needle biopsy using intermittent CT guidance also causes additional radiation exposure for both the patient and the operator.

The performance of ENB in combination with ROSE and PET-CT has not been examined in previous studies. Besides anatomical data and CT scans convertible into multiplanar images using three-dimensional virtual bronchoscopy reconstruction, PET-CT provides functional information on tissue activity and is a discriminator of disease load [18]. There is evidence that the maximum standard uptake value (maxSUV) on PET-CT is an independent predictor of stage and tumor characteristics in non-small-cell lung cancer. The integrated PET-CT has been shown to improve the diagnostic accuracy of the staging of non-small-cell lung cancer [19]. While PET-CT adds valuable information to the staging process, it does not provide a definite tissue diagnosis. In the guidelines for staging non-small-cell lung cancer, the ACCP recommends a PET scan. However, because positive findings of PET scans can result from nonmalignant conditions (e.g., infections), tissue sampling is still required to confirm the suspected malignancy [20].

Rapid on-site cytopathologic examination has been shown to be highly useful and cost effective. It has been shown to improve diagnostic efficacy independent of the localization and histology of the lesion and experience of the operator [21]. While the electromagnetic navigation system helps to reach the target lesion, ROSE ensures that the collected specimen is diagnostic. The combination of ENB and ROSE helps to overcome the limitation of inadequate specimen collection. On the one hand, on-site determination of cytologic adequacy allows the operator to terminate the bronchoscopy and thus minimize the risk of potential complications. On the other hand, on-site determination of cytologic adequacy requires additional collection of specimens to improve the diagnostic yield. A recent study has shown that 65% of patients received a definite malignant or plausible nonmalignant diagnosis when ENB was performed in combination with ROSE [22].

The combination of PET-CT, ENB, and ROSE is a costly procedure. However, there are several indications and subsets of patients in which this procedure appears to be a specifically reasonable workup: (1) the medically inoperable patient requiring tissue confirmation of malignancy before the initiation of radiotherapy with curative intent; (2) the patient with a pulmonary lesion suspicious for a specific benign disease, especially if accompanied by negative or inconclusive PET finding; (3) and the patient preferring confirmation of the diagnosis of a malignant disease before the initiation of more invasive procedures or therapies (e.g., resection).

A limitation of this study is the small number of patients, which is the result of the selection of patients with endobronchially invisible peripheral lung lesions. Another limitation might be the fact that all bronchoscopic procedures were done by the same operator and that the number of biopsy attempts and techniques varied. Therefore, the diagnostic yield of ENB might be influenced by the routine and technique of the bronchoscopist. Factors like lesion size, bronchoscopy technique (rigid versus flexible), conscious sedation versus general anesthesia, availability and use of ROSE, and the expertise of the bronchoscopist will always influence the reported diagnostic yield of a novel method.

ROSE has been shown to be helpful in obtaining diagnostic specimens, and PET-CT can indicate if a negative biopsy might be a false-negative result. Overall, the findings of this study suggest that the combination of ENB, ROSE, and PET-CT greatly increases the diagnostic yield of bronchoscopy and provides a feasible routine workup of peripheral lung lesions.

Copyright information

© Springer Science+Business Media, LLC 2008

Authors and Affiliations

  • Bernd Lamprecht
    • 1
  • Peter Porsch
    • 1
  • Christian Pirich
    • 2
  • Michael Studnicka
    • 1
  1. 1.Department of Pulmonary MedicineParacelsus Medical University HospitalSalzburgAustria
  2. 2.Department of Nuclear Medicine and EndocrinologyParacelsus Medical University HospitalSalzburgAustria

Personalised recommendations