Preoperative Flexible Bronchoscopy in Patients with Persistent Ground-Glass Nodule

There are no accurate data on the diagnostic value of preoperative flexible bronchoscopy (FB) for persistent ground-glass nodule (GGN) of the lung. We evaluated the value of preoperative FB in patients with suspected GGN-type lung cancer. We retrospectively searched a database for subjects who had ‘ground-glass opacity’, ‘non-solid nodule’, ‘part-solid nodule’, or ‘sub-solid nodule’ on chest computed tomography reports between February 2004 and March 2012. Patients who had infiltrative ground-glass opacity lesions, mediastinal lymphadenopathy, or pleural effusion, focal ground-glass opacity lesions >3 cm, and were lost to follow-up were excluded. We assessed the diagnostic value of preoperative FB in patients with persistent GGNs who underwent surgical resection. In total, 296 GGNs were evaluated by FB in 264 patients with persistent GGNs who underwent preoperative FB and surgical resection. The median size of the GGNs was 18 mm; 135 (46%) were pure GGN and 161 (54%) were part-solid GGN. No visible tumor or unsuspected endobronchial metastasis was identified by preoperative FB. Only 3 (1%, 3/208) GGNs were identified preoperatively as malignant by bronchial washing cytology; all were part-solid GGNs. No other etiology was identified by FB. Of the GGNs, 271 (91%) were subsequently confirmed as malignant and 25 (9%) were confirmed as benign at surgical resection. Consequently, the overall diagnostic sensitivity and negative predictive value of preoperative FB on a per-nodule basis was 1% (3/271) and 8% (25/293), respectively. The preoperative FB did not change the surgical strategy. Preoperative FB did not add much to the evaluation of persistent GGNs of the lung. Routine preoperative FB may have limited value in surgical candidates with small persistent pure GGNs.


Introduction
Pulmonary nodules are a common, worrying clinical problem because they often indicate early-stage lung cancer [1,2]. As a low-dose chest computed tomography (CT) screening trial showed a 20% mortality reduction, the management of pulmonary nodules has become an important issue for clinicians with the increased use of chest CT [3]. Recent guidelines for the management of pulmonary nodules have recommended strategies that include observation with serial radiographs, bronchoscopic or transthoracic needle biopsies, and surgical resection [4]. However, choosing the appropriate strategy is difficult because several factors can influence the diagnostic values of the strategies [5].
Persistent ground-glass nodules (GGNs) can indicate focal fibrosis, premalignant lesions, or subtypes of adenocarcinoma [6][7][8], and an early confirmatory diagnosis without surgery is difficult [9]. Follow-up for more than 2 years is needed for GGN-type lung adenocarcinoma due to the long doubling time [10,11]. A bronchoscopic biopsy is usually impossible for GGN-type lung adenocarcinoma because endobronchial metastasis is uncommon and most lesions arise in peripheral areas [12][13][14]. Non-diagnostic transthoracic needle biopsy results cannot rule out the possibility of a malignancy [4]. Therefore, surgical resection with diagnostic and curative intent is usually performed for persistent GGNs that suggest malignancy.
Before the surgical resection of non-small cell lung cancer (NSCLC), flexible bronchoscopy (FB) is performed to identify the underlying etiology, unsuspected endobronchial involvement, and anatomical variation. However, the value of FB in the routine preoperative work-up of pulmonary nodules is still controversial. The American College of Chest Physicians guidelines do not recommend routine preoperative FB for indeterminate small pulmonary nodules [4,15,16], however, recent studies [17,18] have demonstrated the usefulness of preoperative FB, including the ability of FB to identify unsuspected endobronchial involvement and change the planned surgical approach.
These contradictory reports can be explained in part by the fact that several factors influence the diagnostic yield of FB, including the size and location of the nodules and the prevalence of malignancy [4,9,13,15]. However, few reports on the diagnostic value of preoperative FB for pulmonary nodules focus on the radiological characteristics of the nodules [19,20]. Therefore, this study retrospectively evaluated the value of preoperative FB in patients with suspected GGN-type lung cancer who underwent planned surgical resection with diagnostic and curative therapeutic intent.

Study subjects and data collection
We searched a database for subjects treated at Samsung Medical Center, a 1,961-bed referral hospital in Seoul, South Korea, between February 2004 and March 2012 who had 'ground-glass opacity', 'non-solid nodule', 'part-solid nodule', or 'sub-solid nodule' in their chest CT report and reviewed their medical records.
Data on patient characteristics, chest CT and positron emission tomography (PET)/CT findings of GGNs, preoperative FB findings, results of preoperative diagnostic evaluations (bronchial washing cytology, transbronchial lung biopsy, and transthoracic needle biopsy), histopathology, and patient outcomes were collected. Follow-up data were last obtained on December 31, 2012.
This study was approved by the Institutional Review Board of Samsung Medical Center (IRB No. 2012-04-059). The requirement for informed consent from the individual patients was waived given the retrospective nature of the study.

Radiological evaluation of GGNs
A GGN was defined as a rounded area of homogeneous or heterogeneous increased attenuation on CT with a lower density than surrounding soft-tissue structures, such as vessels [21]. The GGNs were subclassified as either pure or part-solid GGN [7,21,22]. Both mediastinal (width, 400 Hounsfield units [HU]; level, 20 HU) and lung (width, 1,500 HU; level, -700 HU) window CT images of the GGNs were viewed and the definitions of pure and part-solid GGNs were based on the tumor shadow disappearance rate (TDR): solid (TDR = 1), part-solid (0 < TDR < 1), and pure (TDR = 0) GGN [23]. Nodules were considered peripherally distributed if they were located in the lateral two-thirds of the lung parenchyma on transverse CT. The number, size (diameter), characteristics (pure or part-solid), margin (smooth or spiculated/lobulated), distribution (central or peripheral), location, and other abnormal findings of the GGNs were evaluated.

Preoperative FB and diagnostic evaluation
A preoperative FB (EVIS BF 1T240 and 1T260; Olympus; Tokyo, Japan) evaluation of the bronchial tree was performed by experienced pulmonologists. All procedures were performed under conscious sedation with midazolam. Local anesthesia was achieved by nebulization with 4% lidocaine. When no endobronchial lesion was visible through the bronchoscope, bronchial washing was performed in the corresponding segmental bronchus. The bronchial washing was performed using 10-20 mL of saline; when the volume of fluid recovered was insufficient, an additional 10-mL saline was injected. The washing aspirates were sent for cytological examination and staining or microbiological culture for bacteria and fungi. A transbronchial lung biopsy or bronchoscopic mucosal biopsy was performed when there was a suspicious lesion, and a transthoracic needle biopsy was performed at the physician's discretion.
All samples were categorized based on the cyto-histopathological reports. The presence of frank malignant cells or rare cells suspicious for malignancy was considered positive. The absence of tumor cells and dysplastic bronchial epithelium and the presence of normal bronchial epithelial cells was considered negative. Surgically confirmed adenocarcinoma was classified according to the new lung adenocarcinoma classification of the European Respiratory Society / International Association for the Study of Lung Cancer / American Thoracic Society [24]. All specimens were evaluated by an experienced lung pathologist (JH). The gold standard for the final diagnosis of the FB was based on the pathology of the surgical specimen.

Statistical analysis
All data are presented as medians (interquartile range [IQR]) for continuous variables and numbers (percentage) for categorical variables. The diagnostic sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of preoperative FB were calculated on a per-nodule basis. Data were compared using the Mann-Whitney U-test for continuous variables and the chi-square or Fisher's exact test for categorical variables. All statistical analyses were performed using PASW 18.0 (SPSS, Chicago, IL) and a two sided p < 0.05 was considered to indicate statistical significance.

Characteristics of the study patients
The database search for subjects who had 'ground-glass opacity', 'non-solid nodule', 'part-solid nodule', or 'sub-solid nodule' on chest CT identified 984 patients. From these, we excluded patients with infiltrative ground-glass opacity lesions, mediastinal lymphadenopathy, or pleural effusion (n = 251), focal ground-glass opacity lesions > 3 cm (n = 18), and those lost to followup (n = 73). Consequently, 642 patients with focal GGNs were identified. Of these 642, patients with transient GGNs (n = 74), patients who did not undergo preoperative FB (n = 136), and patients on observation with CT follow-up (n = 168) were excluded. Ultimately, 264 patients with persistent GGNs who underwent preoperative FB and surgical resection were included in the study (Fig. 1).

Results of diagnostic evaluations
The results of the diagnostic evaluations are summarized according to the characteristics of the GGNs in Table 3. Of the GGNs, 135 (46%) pure GGNs and 161 (54%) part-solid GGNs were evaluated preoperatively by FB. The median size of the part-solid GGNs was significantly larger than the pure GGNs (p < 0.001). Of the 135 pure GGNs, 89% (n = 120) were malignant, adenocarcinoma in situ being the most common (n = 57, 42%), and of the 161 part-solid GGNs, 94% (n = 151) were malignant, invasive adenocarcinoma being the most common (n = 129, 80%). Overall, no visible tumor or unsuspected endobronchial metastasis was identified by preoperative FB, and only a few benign lesions were observed (13 anthracofibrosis, four narrowings or strictures, three mucosal irregularities, and three nodular lesions). Bronchial washing cytology was performed in 208 of the 296 GGNs: 122 part-solid GGN and 86 pure GGN. Only 3 (1%, 3/208) GGNs were identified as malignant preoperatively, all of which were part-solid GGN; no other accompanying etiology was identified by FB. The results of bronchoscopic mucosal biopsies performed on two nodular lesions and three mucosal irregular lesions were all negative. Transbronchial lung biopsies were available only for three part-solid GGNs and all were negative. Consequently, the overall diagnostic sensitivity, specificity, PPV, and NPV of preoperative FB for GGN identification on a per-nodule basis were 1% (3/271) 100% (25/25), 100% (3/3), and 8% (25/293), respectively. Routine preoperative FB did not change the planned surgical strategy in any patient. Table 4 presents the characteristics of the patients who were diagnosed with malignancy by FB preoperatively. All three patients had part-solid GGNs larger than 20 mm with speculated or lobulated margins, and an air bronchogram (n = 2) or bubble lucency (n = 1) was observed. However, there were no unsuspected endobronchial lesions on FB.

Discussion
In this study, no visible tumor or unsuspected endobronchial metastasis was identified by preoperative FB in patients with persistent GGNs. Only three part-solid GGNs were identified preoperatively as malignant by bronchial washing cytology, despite the high incidence of malignancy (271/296, 92%); no other accompanying etiology was identified. Consequently, the overall diagnostic sensitivity and NPV of preoperative FB for GGN identification on a pernodule basis were only 1% (3/271) and 8% (25/293), respectively, and preoperative FB did not change the planned surgical strategy in any patient. Moreover, given that there was no unsuspected endobronchial metastasis and no need to change the surgical approach in 136 patients with persistent GGNs who underwent surgical resection without preoperative FB, in whom 123 (90%) had NSCLC IA, 2 (2%) had NSCLC IB, and 11 (8%) had benign lung diseases, our data suggest that routine preoperative FB may have limited value in surgical candidates with small persistent pure GGNs. These results are supported by previous reports that persistent GGNs occasionally indicate benign lung diseases [8] and GGN-type lung cancer shows minimal invasive growth with relatively long doubling times [10] and a better prognosis than the solid type [10,11,14,[25][26][27]. For example, in a study that evaluated 61 patients with primary lung cancer who underwent 3-year mass CT screening, Hasegawa et al. [10] reported that atypical adenomatous hyperplasia and GGN-type adenocarcinomas had greater volume-doubling times than squamous cell carcinomas. Asamura et al. [27] compared the characteristics of 28 GGN and 20 solid-type lung cancer cases and found no nodal metastasis in the GGN-type lung cancers in contrast to the solid-type cancers, and the 5-year disease-free survival rate was 100% for GGN-type lung cancer. In this context, our data imply that the diagnostic value of FB for persistent lung nodules is dependent on the characteristics of the nodule, not merely its size or location [4,9,13,15], and diagnostic strategies might be adjusted according to the presence of a ground-glass opacity. Other studies of the diagnostic yield of FB for lung cancer have reported that the sensitivity of bronchoalveolar lavage or washing cytology was 29-78% for central lung lesions and 12-65% for peripheral lung lesions [28], which is considerably higher than in our work; however, those studies did not mention the characteristics of the lesions. This might explain the contradictory results regarding the diagnostic value of FB across studies and suggest that discrimination between GGNs and solid-type nodules is important, since this distinction could eventually affect diagnostic strategies. To date, however, only limited data exist on the value of FB for persistent GGNs compared to solid nodules. Further studies focusing on the characteristics of the nodules are needed.
In this study, we also evaluated difference in the diagnostic value of FB between pure and part-solid GGNs ( Table 3) and found that all three GGNs identified preoperatively as malignant by FB were part-solid GGNs, suggesting a relatively lower diagnostic value of FB for pure GGNs compared to part-solid GGNs. However, there was no significant difference in the results of the diagnostic evaluation, except for nodule size, and these results might have resulted from the fact that the overall diagnostic sensitivity of FB for GGNs was very low in our study and only a small proportion of patients underwent aggressive bronchoscopic procedures, such as a transbronchial lung biopsy or mucosal biopsy.
This study had several limitations. First, because the study patients were selected retrospectively based only on chest CT reports, the data should be interpreted conservatively. Second, not all GGNs were subject to bronchoscopic washing cytology, and only a small proportion of patients underwent aggressive bronchoscopic procedures without using guidance, such as CT or fluoroscopy [29]. Therefore, it is possible that the diagnostic value of FB was underestimated. For more accurate analysis, further prospective studies of the value of preoperative FB using various diagnostic techniques in suspected persistent GGNs may be needed.

Conclusions
Preoperative FB did not add much to the evaluation of persistent GGNs of the lung. Routine preoperative FB may have limited value in surgical candidates with small persistent pure GGNs.