Background

With the recent technological and diagnostic advances, and with the widespread use of computed tomography (CT), nodular ground-glass opacities (nGGOs) are being increasingly detected. This increased detection has challenged the diagnosis and management of nGGOs, which have been addressed extensively in the last ten years. However, most approaches are currently not closely grounded in solid evidence [1]. The most recent National Comprehensive Cancer Network (NCCN) guidelines, version 2.2014 for lung cancer screening recommend nonsurgical biopsy or surgical excision of part-solid nodules measuring more than 8 mm that are suspicious for lung cancer upon fluorodeoxyglucose-positron emission tomography (FDG-PET)/CT [2]. For pure GGOs measuring more than 10 mm with stable features upon low-dose CT (LDCT) follow-up, the NCCN guidelines recommend follow-up with LDCT for 6–12 months, nonsurgical biopsy, or surgical excision. On the other hand, in patients with part-solid nodules measuring more than 8 mm, the American College of Chest Physicians (ACCP) guidelines suggest repeat chest CT at 3 months, followed by further evaluation with FDG-PET/CT, nonsurgical biopsy, and/or surgical resection [3]. Especially, in part-solid nodules measuring more than 15 mm, they recommend prompt further evaluation with FDG-PET/CT, nonsurgical biopsy, and/or surgical resection. However, there are currently no clear criteria for deciding nonsurgical biopsy or surgical resection.

Bronchoscopic examination including biopsy is rarely helpful in the diagnosis of nGGOs, as these are usually located peripherally. On the other hand, percutaneous needle aspiration or biopsy (PCNA or PCNB) is an established diagnostic technique that has a high diagnostic yield [4, 5]; however, PCNA or PCNB has a lower sensitivity in smaller nodules [6, 7], and its role in the diagnosis of nGGOs is controversial, with no consensus existing regarding the optimal size threshold or technique [8]. Although recent studies have reported that PCNB provided a high diagnostic accuracy of up to 95% for nGGOs [9, 10], those rates may not be reproducible when strict standards are applied to the biopsy specimens under the new classification of the International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society (IASLC/ATS/ERS), which introduced the concept of adenocarcinoma in situ (AIS) and minimally invasive adenocarcinoma (MIA) [11]. Small biopsy may not be suitable for determining tissue invasiveness.

In addition to pneumothorax or hemoptysis, the risk of malignant cell spread through the tract upon PCNA or PCNB has been addressed [12–15]. Furthermore, the relatively high radiation exposure to the operators is a concern, as percutaneous biopsy of nGGOs is associated with a longer procedure time, owing to the smaller sizes and less solid components of these tumors.

Thin-section CT findings correlate closely with the pathologic diagnosis of nGGOs [16], and the attenuation, marginal characteristics, size, and development of a solid component are potentially helpful to predict malignancy [1, 17, 18]. Therefore, surgical resection of nGGOs without preoperative tissue diagnosis could be a reasonable strategy. However, the necessity of preoperative histologic confirmation by PCNA or PCNB in nGGOs has not been well evaluated.

Seoul National University Bundang Hospital (SNUBH) has used its own protocol for the management of nGGOs (Figure 1) that is basically similar to the NCCN guideline [2]. Briefly, we suggest direct surgical resection of nGGOs highly suspicious for malignancy, rather than nonsurgical biopsy. In this study, we evaluated the rate of malignancy, complications related to surgery, and the cost benefits of resecting nGGOs without prior tissue diagnosis when those nGGOs were highly suspected for malignancy based on their size, radiologic characteristics, and clinical courses.

Figure 1
figure 1

Seoul National University Bundang Hospital guidelines for pure GGO (A) and part-solid GGO (B). GGO, ground-glass opacity; LDCT, low-dose computed tomography; HRCT, high resolution computed tomography.

Full size image

Methods

Study population and study design

We retrospectively reviewed the medical records of all patients who underwent surgical resection of nGGOs between January 2009 and October 2013 at SNUBH in Seoul, Korea. Patients who underwent preoperative tissue diagnosis by PCNA or PCNB were excluded. The primary outcomes were the rate of malignancy of nGGOs among patients who underwent surgical resection without prior tissue diagnosis, and the clinical and radiological predictors of malignancy. The secondary outcomes were complications related to surgery and differences in the total costs, days of hospitalization, and the time interval before surgery between patients with or without nonsurgical histologic diagnosis. This study was approved by the institutional review board of SNUBH (L-2014-274). The requirement for informed consent was waived.

Radiologic evaluation

CT scans were obtained using various instruments, including the Brilliance-64, MX-8000 IDT, and iCT 256 (Philips Medical Systems, Cleveland, OH, USA). Imaging was obtained using a lung window setting with a level of -600 Hounsfield units (HU) and a width of 1500 HU, and a mediastinal window setting with a level of 30 HU and a width of 400 HU. Scanning was performed from the thoracic inlet to the upper portion of the kidneys. When there were multiple nGGOs in a patient, only nGGOs with permanent pathologic confirmation were selected based on the surgical records. The nGGO lesions containing patchy opacities that totally obscured the lung parenchyma were classified as part-solid GGOs, whereas if no part of the encircled lung parenchyma was completely obscured, they were classified as pure GGOs [19]. In each nGGO, the presence of a solid component, an air-bronchogram, bubble lucency, pleural or fissure retraction, and margin irregularity were evaluated. The maximal diameters and tumor disappearance rate were also measured [20]. In 191 nGGOs of 173 patients who underwent FDG-PET/CT, data on the maximal standardized uptake values (SUVs) were also collected.

Serial CT scans performed at least every 4 weeks were available for 244 nGGOs in 223 patients over a median follow-up duration of 9.1 months (range, 7.3-123.9 months). The interval change was investigated in these nGGOs, and progression of nGGO was defined as (1) ≥2 mm increase in the GGO size, (2) ≥2 mm increase in the solid component, or (3) emerging new solid component of any size [21, 22].

Total costs, days of hospitalization, and waiting time

Total costs were calculated as the sum of the costs for both the diagnosis and treatment of nGGOs during hospitalization. We excluded all charges incurred in the outpatient clinic. Costs were converted to US dollars according to the current average exchange rate (1 US dollar =1025 won). Days of hospitalization were defined as the total length of hospital stay for both surgical resection and work-up. Waiting time (the time interval before surgery) was defined as the interval between the first hospital day of admission for work-up and the day of operation, even if a patient was discharged and readmitted for surgery [23].

Statistical analysis

Continuous data are presented as mean ± standard deviation (SD), whereas categorical data are presented as numbers and percentages. The relationships between the clinical and radiological characteristics and the final pathologic diagnosis were evaluated using the independent-sample t-test for continuous variables and the χ2 test or Fisher’s exact test for categorical variables. Univariate and multivariate analyses were performed and the results were described with odds ratio (OR) and 95% confidence interval (CI). P values less than 0.05 were considered to have statistical significance. All statistical analyses were performed using SPSS for Windows, version 19.0 (SPSS Inc., Chicago, IL, USA).

Results

Patient characteristics

Of 356 nGGOs in 324 patients who underwent surgical resection for nGGOs from January 2009 and October 2013, 330 nGGOs (92.7%) in 300 patients were resected without preoperative tissue diagnosis (Figure 2). The main indications for direct surgical resection were a tumor size more than 10 mm (n =291, 88.2%), morphologic characteristics (an air-bronchogram, bubble lucency, pleural or fissure retraction, or irregular margin) on CT scans (n =255, 77.3%), and an increase ≥2 mm in the whole GGO size (n =87, 26.4%). All other reasons that are not indicated by our protocol are listed in Figure 2. Among 21 nGGOs, six nGGOs were co-resected with indicated GGO nodules and one nGGO was resected because metastasis of underlying thyroid cancer was strongly suspected. The decisions to resect other nGGOs that did not meet the SNUBH protocol were mainly influenced by the patients’ will.

Figure 2
figure 2

Flow chart. Of the 330 resected nGGOs, 314 were diagnosed as lung adenocarcinoma, and the rate of malignancy was 95.2%. aAmong 21 nGGOs, six nGGOs were co-resected with indicated GGO nodules and one nGGO was resected because metastasis of underlying thyroid cancer was strongly suspected. The decision to resect the other 14 GGOs was mainly influenced by the patients’ will. nGGOs, nodular ground-glass opacities; AIS, adenocarcinoma in situ; MIA, minimally invasive adenocarcinoma; ADC, adenocarcinoma; AAH, atypical adenomatous hyperplasia.

Full size image

Of the 330 resected nGGOs, 314 were diagnosed as lung adenocarcinoma, and the rate of malignancy was 95.2%. A total of 242/255 (94.9%) part-solid GGOs, and 72/75 (96.0%) pure GGOs were malignant lesions. Only 16 nGGOs (4.8%) were proven to be benign lesions, including atypical adenomatous hyperplasia (AAH). Regarding the demographic characteristics, there was no significant differences in either age (p =0.127) or sex (p =0.461) between patients with malignant and benign lesions (Table 1). Moreover, smoking history (p =0.555) did also not differ between these groups.

Table 1 Baseline and radiologic characteristics of the study patients (n =330)
Full size table

Radiologic characteristics

The mean values of the maximal diameter ± SD of the nGGOs were 15.1 ± 9.3 mm and 20.3 ± 11.0 mm for benign and malignant lesions, respectively (p =0.063) (Table 1). The morphologic characteristics of the nGGOs in terms of the presence of a solid component, an air-bronchogram, bubble lucency, pleural or fissure retraction, and margin irregularity did not significantly differ between the groups. Moreover, for the 191 nGGOs with FDG-PET/CT findings, the maximal SUV was not significantly different between benign and malignant lesions.

Of 244 nGGOs followed-up over 9.1 months (range, 7.3-123.9 months), 92 (37.7%) showed tumor progression. However, progression of nGGOs was not associated with the risk of malignancy. Of 12 benign nGGOs, four (33.3%) showed progression. Three nGGOs displayed increases in the whole GGO size, and the one pure GGO had become a part-solid GGO. Two of the three nGGOs displaying increases in the whole GGO size were subpleural fibrosis, whereas the other one was diagnosed as AAH on final pathology. The one nGGO becoming a part-solid nodule turned out to be an anthracofibrotic nodule. On the other hand, one nGGO had decreased in size, which was an adenocarcinoma, acinar predominant.

In the multivariate analysis, larger size was the only identified independent factor that was predictive of malignancy (OR, 1.086; 95% CI, 1.001-1.178; p =0.047). Especially, nGGOs more than 15 mm in size were significantly associated with a higher risk of malignancy compared to nGGOs less than 10 mm in the multivariate analysis (OR, 8.323; 95% CI, 1.968 – 35.196; p =0.004) (Table 2).

Table 2 Multivariate analysis for the risk of malignancy according to ground glass opacity size
Full size table

Surgical procedure and complications related to surgery

Limited surgical resection was conducted in 170/330 (51.5%) nGGOs. 324 nGGOs (98.2%) were resected by video-assisted thoracoscopic surgery (VATS), and conversion of VATS to open thoracotomy occurred in only 4 cases (1.2%). Complications related to surgery occurred in 6.7% (22/330) of cases. The most common complications were prolonged air leak for more than 7 days (n =15; 68.2%), pleural effusion (n =4; 18.2%), pneumothorax (n =2; 9.1%), and chylothorax (n =1; 4.5%). All 16 nGGOs diagnosed as benign lesions were resected by VATS, and only one of these patients experienced postoperative complications (prolonged air leak).

Pathologic diagnosis

Table 3 shows the pathologic diagnoses of the patients. Of the 314 malignant nGGOs, 38 (12.1%) were diagnosed as AIS and 63 (20.1%) as MIA and 213 (67.8%) were diagnosed as invasive adenocarcinomas. Of 213 (67.8%) invasive adenocarcinomas, 115 (36.6%) were acinar predominant, 52 (16.6%) were papillary predominant, and 33 (10.5%) were lepidic predominant. The most common pathologic findings in the 16 benign nGGOs were focal interstitial fibrosis (n =5; 31.3%), AAH (n =4; 25.0%), and subpleural fibrosis (n =3; 18.8%).

Table 3 Pathologic diagnoses
Full size table

Advantages on costs, hospital stay, and waiting time

Next, we compared the total costs, days of hospitalization, and the waiting times between patients who underwent surgical resection of nGGOs with or without preoperative tissue diagnosis. Surgical resection without histologic confirmation was found to be associated with significant decreases in the total costs, hospital stay, and waiting times (Table 4). Approximately 2,546 US dollars were saved by performing surgical resection without tissue diagnosis as compared to with tissue diagnosis (9,271 vs. 11,817 US dollars, p =0.004). Moreover, the mean hospital stay was 3.0 days shorter (6.8 vs. 9.8 days, p =0.015) and the mean waiting time was 6.3 days shorter (2.5 vs. 8.8 days, p =0.001) in patients who underwent surgical resection without tissue diagnosis.

Table 4 Hospital stays, waiting times, and costs for patients undergoing surgical resection with and without tissue diagnosis
Full size table

Discussion

The aim of this retrospective study was to investigate the necessity of preoperative biopsy for GGO nodules which were suspicious for malignancy. We evaluated the rate of malignancy, complications related to surgery, and the cost benefits of surgical resection of nGGOs without preoperative tissue diagnosis when those nGGOs were highly suspicious for malignancy in terms of size, radiologic characteristics, and clinical courses. Lack of adequate control group, such as randomly assigned GGO nodules with preoperative biopsy was the main limitation of this study. However, our study yielded four main findings: (1) the rate of malignancy was high (95.2%) in nGGOs highly suspected for malignancy based on clinical and radiologic characteristics; (2) only tumor size was a significant independent predictor of malignancy in the multivariate analysis; (3) the rate of complications related to surgery was low (6.7%), with no mortality and minimal morbidity; and (4) direct surgical resection without tissue diagnosis significantly reduced the total costs, days of hospitalization, and waiting time to surgery.

The role of PCNA or PCNB in the diagnosis of nGGOs remains limited [8]. Hur et al. [24] reported the sensitivity of CT fluoroscopy-guided needle biopsy as 67% for diagnosing malignancy in 28 patients with nGGO lesions. In another study of 40 individuals with nGGOs, the diagnostic yield of percutaneous CT-guided core biopsy was 84% (16/19), whereas it was non-diagnostic in three patients (16%) [25]. Two of these underwent surgical resection and were diagnosed as lung adenocarcinoma. In our study, 26 patients underwent PCNA or PCNB and the diagnostic accuracy was found to be 65.4% (17/26) under the IASLC/ATS/ERS classification of lung adenocarcinoma. All nGGOs that were resected after PCNA or PCNB were in fact malignant. Four nGGOs diagnosed as benign lesions and another four nGGOs with non-diagnostic results on PCNA or PCNB underwent surgical resection based on strong clinical suspicion for lung cancer, and were demonstrated to be invasive adenocarcinomas.

Recently, several studies have shown that the diagnostic accuracy of PCNB was greater than 90% for nGGOs [9, 10]; however, these results may be dependent on the experience and skills of the operators, and may hence not be fully reproducible. In this context, the third edition of the ACCP guidelines on the diagnosis and management of lung cancer stated that nonsurgical biopsy should not be used to exclude malignancy considering its unsatisfactory sensitivity and limited negative predictive value [3].

VATS lobectomy for patients with early-stage lung cancer is a standard surgical treatment, and is associated with lower morbidity and improved survival rates compared with open thoracotomy [26]. Recently, several studies have suggested that thoracoscopic limited resection is a valid surgical technique for nGGOs selected by thin-section CT scans [27]. With the widespread use of VATS, it is possible to diagnose and treat nGGOs simultaneously.Recommendations for the management of nGGOs have been used by SNUBH (Figure 1). These recommendations state that, regardless of the presence of a solid component, surgical resection should be considered if there is an increase in size ≥2 mm or development of a solid component in a pure GGO. In pure GGOs ≥10 mm, we suggest repeat chest CT at 3 months. In GGOs without significant changes in the initial 3 months of follow-up, we recommend surgical excision for nGGOs ≥15 mm, whereas we recommend chest CT follow-up for one year or surgical excision for nGGOs measuring 10–15 mm in size. In part-solid GGOs ≥10 mm with clinical suspicion of malignancy, we recommend surgical resection even if these do not show significant changes at the initial one-month follow-up.

Heo et al. [23] evaluated 113 patients who underwent surgical resection in SNUBH from January 2008 to May 2009 without prior tissue diagnosis for highly suspicious pulmonary nodules, including solid and GGO lesions. In their retrospective study, 45/50 (90%) patients with nGGOs had malignancy; and they reported that presence of a solid component, bubble lucency, irregular margin, and larger size correlated with malignancy. Although many other studies have also reported that larger size, irregular border, partly solid attenuation, internal air bronchograms, and central bubbly lucency were associated with higher rates of malignancy [19, 28, 29], some studies have reported conflicting results. In a study of 53 pure nGGOs in 49 patients, no significant differences in the morphologic characteristics and size were observed between malignant and benign nodules [30], although it is possible that this study was underpowered to detect differences [3]. However, in our study including a larger number of nGGOs, the maximal diameter was the only predictive factor of malignancy, and there were no significant differences in the morphologic features on CT between malignant and benign lesions.

We speculate that the failure to detect morphologic factors to distinguish benign from malignant nGGOs was mainly influenced by selection bias. Not all patients with an nGGO lesion underwent surgical resection. Only patients who were highly suspected to have malignancy based on the tumor size, radiologic characteristics, and clinical courses underwent surgical resection, although 14 of the resected nGGOs did not meet the SNUBH protocol in this study. This may explain the higher malignancy rate (95.2%) and lower proportion of AAH (4/330, 1.2%) in the present study compared to previous studies, which have reported malignancy rates between 58.7-75.0% [19, 30] and proportions of AAH between 5.7%-20.9% [30, 31].

Conclusions

In conclusion, upon careful selection of nGGOs that are highly suspicious for malignancy, surgical resection of nGGOs without tissue diagnosis is recommended as it reduces costs and the length of hospital stays.