Prognostic influence of a ground-glass opacity component in hypermetabolic lung adenocarcinoma

Abstract

 

Open in new tabDownload slide

OBJECTIVES

We evaluated the oncological role of a ground-glass opacity (GGO) component in hypermetabolic lung adenocarcinoma with a high maximum standardized uptake value.

METHODS

Between 2008 and 2017, we retrospectively reviewed the data of surgically resected clinically node-negative lung adenocarcinomas with a hypermetabolic activity. Furthermore, they were classified based on the presence of GGO. The prognostic significance of a GGO in hypermetabolic tumours was evaluated using the Cox proportional-hazards model. The overall survival (OS) was estimated by the Kaplan–Meier method using a log-rank test.

RESULTS

Of the 1134 surgically resected clinically node-negative lung adenocarcinoma, 603 cases with hypermetabolic activity (maximum standardized uptake value ≥3.0 mg/dl) were evaluated. Among them, there were 120 (20%) cases in the With GGO group and 483 (80%) in the Solid group. The 5-year OS of patients with hypermetabolic lung adenocarcinoma was significantly different between the With GGO and Solid groups (83.1% vs 59.4%, P < 0.001). After stratifying by the presence of GGO, maximum standardized uptake value and maximum tumour size were independently significant prognosticators in the Solid group by multivariable analysis; however, no clinical factors were associated with survival among the With GGO group. The 5-year OS was favourable despite the solid component size among the With GGO group (T1a + T1b: 85.5%, T1c: 80.0%, T2a or more: 84.2%, P = 0.904). For the Solid group, survival diminished drastically with increasing tumour size (T1a + T1b: 68.7%, T1c: 62.8%, T2a or more: 48.0%, P < 0.001).

CONCLUSIONS

The prognosis of lung adenocarcinoma with GGO was favourable even in hypermetabolic tumours. Accordingly, the presence of GGO should be considered as an important parameter in the next clinical T classification.

INTRODUCTION

With an increasing application of thin-section computed tomography (CT) scan for the diagnosis of lung cancer [1], more and more lung nodules characterized by ground-glass opacity (GGO) are being detected in daily practice. Following the reports of the various studies that have previously elucidated the correlation of malignant behaviour between radiological and pathological findings in lung adenocarcinomas, solid component size and consolidation-to-tumour ratio (CTR) on thin-section CT scan have been considered as important radiological parameters that reflect the prognosis [2–4], and this has impacted the current 8th edition of TNM staging of non-small-cell lung cancer (NSCLC) [5]. However, due to the excellent prognosis of lung cancers with a GGO component compared to those with purely solid nodules of equivalent consolidation size, the presence of a GGO component in itself is considered a novel prognosticator in early-stage lung cancer. To date, this simple notion has been receiving an increased focus as a reliable and favourable prognostic indicator in early-stage NSCLCs [6–12].

In addition, a maximum standardized uptake value (SUVmax) on 18F-fluoro-2-deoxyglucose positron emission tomography (FDG-PET) has been widely recognized as a surrogate of tumour aggressiveness, which is useful in predicting the pathological invasiveness and prognosis in early-stage lung cancer [13–18]. In such circumstances, only a small GGO component could predict better oncological outcomes in resectable lung cancer [7, 10]; however, the oncological role of a GGO component is still controversial, especially in hypermetabolic lung adenocarcinoma showing a high SUVmax value. Therefore, this study aimed to systematically investigate the prognostic influence of a GGO component in resectable hypermetabolic lung adenocarcinoma. We hypothesized that the presence or absence of a GGO component would be considered a strong parameter for a favourable prognosis even in hypermetabolic lung adenocarcinoma, which could be confirmed regardless of the solid component size in tumours with a GGO component.

MATERIALS AND METHODS

Study population

Between 2008 and 2017, we retrospectively reviewed the data of surgically resected clinically node-negative lung adenocarcinomas with a hypermetabolic activity. At our institute, the 8th edition of the TNM classification system is used for clinical staging [5, 19]. There were no missing data for the variables examined in this study, except data of epidermal growth factor receptor gene mutation. The inclusion criteria were preoperative staging determined by thin-section CT and complete resection without preoperative chemo- and/or radiotherapy. With regard to the clinical nodal assessment, clinical-N0 indicated both non-enlarged lymph nodes (short axis <10 mm) on thin-section CT and no uptake on PET scan. Invasive modalities for mediastinal lymph node staging, such as mediastinoscopy or endobronchial ultrasound-guided transbronchial needle aspiration, were not performed preoperatively in clinical N0 patients at our institute. The medical records of each patient were reviewed retrospectively under a waiver of individual informed consent approved by the institutional review board of Juntendo University School of Medicine, Tokyo, Japan (IRB: 18-319).

Radiological evaluations of thin-section CT scan findings

For all the patients, the findings of the preoperative thin-section CT scan were reviewed in detail by the authors (A.H., T.M. and K.S.) and a radiology oncologist. Tumour size was determined preoperatively based on the thin-section CT findings. In addition, all the tumours were subsequently evaluated to estimate the extent of the GGOs by the thin-section CT scan with maximum slice thickness of 2-mm collimation. With regard to the institutional equipment from 2008 to 2017, 4-detector-row CT scanner or 16-detector-row CT scanner (Aquilion or Aquilion 16; Canon Medical Systems, Tochigi, Japan) was mainly used for the lung cancer diagnosis. The lung was photographed with a window level of −500 to −700 H and a window depth of 1000–2000 H as a ‘lung window’, and a window level of 30–60 H and a window depth of 350–600 H as a ‘mediastinal window’. GGO was defined as an area of a slight, homogenous increase in density that did not obscure the underlying vascular markings. Furthermore, the CTR was defined as the ratio of the maximum size of a consolidation to the maximum tumour size on thin-section CT scan [2]. In this study, all the tumours were sub-classified into 2 groups based on the presence of a GGO component on thin-section CT scan, i.e. With GGO group (0 ≤ CTR < 1.0) or Solid group (CTR = 1.0)⁹.

Pathological evaluations

The lung adenocarcinomas were histologically classified according to the IASLC/ATS/ETS classification as adenocarcinoma in situ, minimally invasive adenocarcinoma and invasive adenocarcinoma [20]. The histological subtypes were classified according to the predominant subtype after comprehensive histological subtyping, which implies a semi-quantitative estimation of the percentage of different subtypes in 5% increments. In the current study, lepidic predominant adenocarcinoma (LPA) was defined as a tumour that showed the lepidic component most frequently among the invasive adenocarcinomas. Furthermore, pathological invasive lung adenocarcinoma was defined as a tumour with at least one of the following pathological factors: nodal metastasis, lymphatic invasion or vascular invasion based on the JCOG0201 study in Japan [2].

Operation policy

With regard to the operative modes at our institution, if a tumour is GGO dominant or purely GGO, the patient is a candidate of limited surgical resection [21]. While major lung dissection with systematic or selective lymph node dissection is warranted for a part-solid or solid tumour, intentional segmentectomy with systematic or selective lymph node dissection is now indicated for part-solid or solid lung cancers that are 2 cm or smaller in size following a prospective randomized trial [22]. Non-anatomic wedge resection was performed for a few patients who were elderly or had a high cardiopulmonary risk.

Follow-up policy

The routine follow-up evaluation included a physical examination, chest radiography, chest CT scan and blood tests including measurements of tumour markers every 6 to 12 months. If any symptom or sign of recurrence was observed, further evaluation was performed, including CT, brain magnetic resonance imaging and PET/CT to assess the loco-regional or distant cancer recurrence. Locoregional recurrence was defined as occurrence within the residual same lobe and hilum or mediastinal lymph nodes, which was essentially diagnosed by cytological or histological confirmation based on biopsy or surgical resection.

Statistics

In this study, we initially investigated the prognostic influence of a GGO component in the tumros with a hypermetabolic activity. Subsequently, the prognostic significance of GGO in the hypermetabolic lung adenocaricnoma was evaluated using the Kaplan–Meier method and Cox regression for the entire cohort. Furthermore, to investigate the clinicopathological features or survival outcomes based on the presence of GGO component, subgroup analyses were exploratory implemented in both the entire cohort and each clinical T category. The continuous variables that were normally distributed were summarized by their mean and standard deviation; those not normally distributed were summarized by their median and interquartile range. The chi-squared test was used to compare the factors. The unpaired t-test was used if the continuous variables were normally distributed, and if not, the Wilcoxon rank-sum test was used. The Cox proportional-hazards model was fit to adjust the effect of GGO and to identify the significant prognostic factors for the overall survival (OS) using SPSS Statistics 27 (IBM Inc., USA). Forward and backward stepwise procedures were used to determine the combination of factors essential for the survival outcomes. The OS was estimated using the Kaplan–Meier method and compared by the log-rank test across the different groups. The date of surgical resection was set as the starting point and the date of death or last follow-up as the end point of OS. Hypermetabolic lung adenocarcinoma was defined with an SUVmax value to predict pathological invasive lung adenocarcinoma [2] using a receiver operating characteristic (ROC) curve. Regarding the way to select the optimal cut-off value from the ROC curve, we calculated the distance between the point (0, 1) and each observed cut-off point on the ROC curve. The optimal cut-off value was obtained from the point at which the distance is minimum. With regard to multiple comparisons in the subgroup analyses, P-value was adjusted by Bonferroni procedure. A difference was considered statistically significant when the P-value was <0.05.

RESULTS

In this study, 1134 surgically resected clinically node-negative lung adenocarcinoma was comprised 525 With GGO tumours and 611 solid tumours. Among them, we defined hypermetabolic lung adenocarcinoma as tumours with SUVmax ≥ 3 g/dl (i.e. high SUVmax), due to the result of an ROC curve to evaluate the association between tumour SUVmax and pathological invasive lung adenocarcinoma (Supplementary Material, Fig. S1). As a result, Of 603 eligible cases with high SUVmax value, 120 (20%) patients showed With GGO tumour and 483 (80%) showed a solid tumour on thin-section CT scan (Fig. 1).

The overall clinicopathological characteristics of hypermetabolic lung adenocarcinomas are summarized in Table 1. Among them, solid component size on thin-section CT scan was significantly larger in the Solid group (P < 0.001). Furthermore, the propotion of histological non-lepidic predominant tumour and pathological nodal metastasis was significantly higher in the Solid group compared to the With GGO group (P < 0.0001). Table 2 presents the results of the Cox proportional-hazards model. On multivariable analysis of hypermetabolic lung adenocarcinoma, the presence of a GGO component was an independent significant prognostic factor of the OS in the entire cohort (hazard ratio 0.316, 95% confidence interval 0.117–0.854, P = 0.023). The 5 year-OS of hypermetabolic lung adenocarcinoma in this study was 63.9%, with a median follow-up time of 54 months. It was significantly different between the With GGO group and the Solid group (Fig. 2A; 83.1% vs 59.4%, log-rank test P < 0.001). Lung cancer recurrence was observed in 189 patients [14 (11.6%) in the With GGO group and 175 (36.2%) in the Solid group, P < 0.001]. Regarding recurrence, 1 local recurrence, 11 distant and 2 local and distant recurrences were detected in the With GGO group, while 42 local recurrences, 89 distant and 44 local and distant recurrences were detected in the Solid group.

After stratifying by the presence of a GGO component, multivariable analyses revealed that the carcinoembryonic antigen level (P = 0.002), SUVmax (P = 0.044), maximum tumour size (P < 0.001), pathological nodal metastasis (P = 0.004) and lymphatic invasion (P = 0.010) were independently significant prognosticators in the Solid group, however, the clinical and pathological factors, excluding gender and nodal involvement, were not associated with the OS among the With GGO group (Table 2). When we evaluated the OS of hypermetabolic lung adenocarcinoma stratified by the presence of a GGO component, the 5-year OS was favourable despite the solid component size in the With GGO group [Fig. 2B; T1a + T1b (n = 67): 85.5%, T1c (n = 34): 80.0%, T2a or more (n = 19): 84.2%, log-rank test P = 0.904]. However, in the Solid group, the 5-year OS split almost fairly with an increasing size of the tumour [Fig. 2C; T1a + T1b (n = 150): 68.7%, T1c (n = 160): 62.8%, T2a or more (n = 173): 48.0%, log-rank test P < 0.001].

Table 3 demonstrates the exprolatory comparison of the clinicopathological characteristics of hypermetabolic lung adenocarcinoma in each clinical T category. When they were stratified based on a GGO component, the solid component size was similar between the With GGO group and the Solid group in T1c/T2a or more tumours. Furthermore, the mean CTR of the With GGO group was 0.79 in the T1c and 0.84 in the T2a or more lesions, which indicated radiologically solid-predominant tumours only with a small amount of GGO component. Pathologically, however, the frequency of non-lepidic adenocarcinoma and pathological lymphatic or vascular invasion and nodal metastasis in the Solid group were significantly more than those of the GGO group in many aspects. Furthermore, the 5-year OS was assessed based on the presence of a GGO component in each clinical T category (Fig. 3), and the With GGO group was associated with better survival compared to the Solid group among the hypermetabolic lung adenocarcinoma showing a high SUVmax value [Fig. 3A (T1a + T1b): 85.5% vs 68.7%, P = 0.039, Fig. 3B (T1c): 80.0% vs 62.8%, P = 0.024, Fig. 3C (T2a or more): 84.2% vs 48.0%, P = 0.023].

DISCUSSION

The prognostic value of a GGO component and its effect on T classification in NSCLC staging is an important clinical issue, and the concept has been validated using nationwide long-term follow-up data in Japan [9]. In the current study, we aimed to evaluate whether these clinical notions concerning the oncological impact of a GGO component are similarly adopted to the tumours with high malignant potential showing an elevated SUVmax value. As a result, a favourable prognosis of lung adenocarcinoma with a GGO component was confirmed even in hypermetabolic diseases, which was regardless of the solid component size among the tumours with a GGO on thin-section CT scan. Accordingly, the presence or absence of a GGO should be strongly considered as an important novel parameter in the next clinical T classification.

Furthermore, this study presented that the prognosis was clearly distinct between tumours with and those without a GGO component among hypermetabolic lung adenocarcinomas in each clinical category. It might be generally expected that the overall excellent prognoses of lung cancer with a GGO component are largely dependent on less-invasive lung adenocarcinomas which are radiologically GGO dominant or have lower SUVmax. Meanwhile, a consensus would not be reached whether the prognostic influence of the presence of a GGO component is applied to those of a high SUVmax value. However, when we excluded tumours featuring a GGO predominance with lower SUVmax values from the entire cohort, the presence or absence of a GGO component was still strongly associated with both the prognostic and clinicopathological distinctions in hypermetabolic lung adenocarcinoma. In contrast, the oncological outcomes of pure-solid tumours are further dismal, and the prognoses were fairly split based on the tumour size. In the current series presenting a high SUVmax, we confirmed differences between the 2 groups regarding the frequency of histological lepidic growth, pathological nodal involvement and survival outcomes, regardless of a similar solid component size being shown. Based on these analyses, lung cancer with a GGO component is considered to be of a different oncological category from that of the purely solid tumours.

Theoretically, these significant differences in the clinicopathological and oncological outcomes would reflect the distinct carcinogenesis of lung adenocarcinoma among the 2 lesions [23–27]. As reported in several retrospective studies, the malignant potential of lung adenocarcinoma with a GGO is inferior to that of pure-solid tumour [6–12, 25, 26]. The differences of epidermal growth factor receptor gene mutation status among the 2 study group might relate to the distinct carcinogenesis [28, 29]. Furthermore, as shown in multivariable analyses of the subgroup, several other biological factors might influence the intrinsic differences in the 2 study groups, not only for the presence of GGO. Meanwhile, tumour SUVmax value has been widely recognized as a surrogate of tumour aggressiveness, which is useful in predicting pathological invasiveness, prognosis, or appropriate operative strategies in early-stage lung cancer [14, 15, 18]. Furthermore, previous reports have indicated that hypermetabolic tumours were associated with a substantially higher incidence of occult pathological nodal metastasis and worse survivals [16, 17]. However, our study revealed a low incidence of pathological nodal metastasis or loco-regional recurrence even in the tumour with high SUVmax provided that the tumour showed a GGO component. If the presence of GGO translates into a less-aggressive biological behaviour, even for hypermetabolic tumour, it would be a rationale to change the current paradigm of surgical indications for sublobar resections.

To date, the efficacy of intentional segmentectomy is under investigation and great caution is required to indicate sublobar resections for radiologically invasive lung cancers from the point of loco-regional cancer control. Hence, we should await the result of randomized phase III trial in Japan and US for peripherally located small-sized lung cancer. Actually, however, a recent study reported that lobectomy and segmentectomy are comparable oncological procedures for patients with carefully staged cT1 lung adenocarcinoma with hypermetabolic tumours (SUVmax ≥3 g/dl) [13]. These backgrounds might indicate that the presence of lung adenocarcinoma with a GGO component is a potential candidate of sublobar resection for peripherally located small-sized lung cancers, despite its hypermetabolic status. In contrast, our previous study demonstrated that the loco-regional cancer control and survival outcomes of sublobar resection were significantly worse compared to those of lobectomy in the high SUVmax tumours with a radiological pure-solid appearance [30]. In general, lobectomy with nodal dissection would be a mainstay surgical procedure from the standpoint of definitive loco-regional management, especially when tumours show a high SUVmax because of their truly highly invasive nature. Among the radiological solid tumours, however, segmentectomy may be an optimal surgical mode in case they show low SUVmax because they revealed similar oncological outcomes as the cases of lobectomy [30], although further validation studies are necessary to confirm the result.

Limitations

There are some limitations to this study. First, this study was limited by its retrospective nature in a single centre. In addition, SUVmax values limit the generalizability of our results in daily practice. There are variations in SUVmax quantification among several institutions, and SUVmax is easily influenced by many factors, such as patient preparation procedures, scan acquisition, image reconstruction and data analysis. To overcome this disadvantage, the standardized uptake value needs to be revised and PET procedures standardized. Therefore, further investigations are warranted in the future. Finally, the prevalence of a GGO component in hypermetabolic lung adenocarcinoma seems to be essentially higher in Asia. We need some sort of validation from other continents to consolidate the prognostic value of the presence of a GGO component in hypermetabolic lung adenocarcinoma.

CONCLUSION

In conclusion, a favourable prognostic impact of the presence of a GGO component was confirmed in hypermetabolic lung adenocarcinoma. Lung cancer with a GGO component is considered to be of a different oncological category from that of solid tumours without a GGO component.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

ACKNOWLEDGEMENTS

The authors truly appreciate Shuko Nojiri, PhD, Medical Technology Innovation Center, Juntendo University, for the support of statistical analysis and Kazuhiro Suzuki, MD, Department of Radiology, Juntendo University, for the support of radiological evaluation of the tumours. This work was supported in part by a Grant-in-Aid for Cancer Research from the Ministry of Health, Labour and Welfare, Japan, and the National Cancer Center Research and Development Fund (26-A-4). We also like to thank Editage (www.editage.com) for English language editing.

Conflict of interest: none declared.

Author contributions

Aritoshi Hattori: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Visualization; Writing—original draft. Takeshi Matsunaga: Conceptualization; Data curation; Methodology; Writing—review & editing. Mariko Fukui: Conceptualization; Data curation; Writing—review & editing. Kazuya Takamochi: Conceptualization; Data curation; Methodology; Supervision; Writing—review & editing. Kenji Suzuki: Conceptualization; Data curation; Funding acquisition; Methodology; Supervision; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Stefano Bongiolatti, Larry R Kaiser and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

 
• CT

  Computed tomography

 
• CTR

  Consolidation-to-tumour ratio

 
• GGO

  Ground-glass opacity

 
• NSCLC

  Non-small-cell lung cancer

 
• OS

  Overall survival

 
• SUVmax

  Maximum standardized uptake value

 
• ROC

  Receiver operating characteristic

Author notes