Lung Adenocarcinomas Manifesting as Radiological Part-Solid Nodules Define a Special Clinical Subtype

Diagnosis of Invasive Lung Adenocarcinoma Based on Chest CT Radiomic Features of Part-Solid Pulmonary Nodules: A Multicenter Study

Background Solid components of part-solid nodules (PSNs) at CT are reflective of invasive adenocarcinoma, but studies describing radiomic features of PSNs and the perinodular region are lacking. Purpose To develop and to validate radiomic signatures diagnosing invasive lung adenocarcinoma in PSNs compared with the Brock, clinical-semantic features, and volumetric models. Materials and Methods This retrospective multicenter study (https://ClinicalTrials.gov, NCT03872362) included 291 patients (median age, 60 years; interquartile range, 55-65 years; 191 women) from January 2013 to October 2017 with 297 PSN lung adenocarcinomas split into training (n = 229) and test (n = 68) data sets. Radiomic features were extracted from the different regions (gross tumor volume [GTV], solid, ground-glass, and perinodular). Random-forest models were trained using clinical-semantic, volumetric, and radiomic features, and an online nodule calculator was used to compute the Brock model. Performances of models were evaluated using standard metrics such as area under the curve (AUC), accuracy, and calibration. The integrated discrimination improvement was applied to assess model performance changes after the addition of perinodular features. Results The radiomics model based on ground-glass and solid features yielded an AUC of 0.98 (95% confidence interval [CI]: 0.96, 1.00) on the test data set, which was significantly higher than the Brock (AUC, 0.83 [95% CI: 0.72, 0.94]; P = .007), clinical-semantic (AUC, 0.90 [95% CI: 0.83, 0.98]; P = .03), volumetric GTV (AUC, 0.87 [95% CI: 0.78, 0.96]; P = .008), and radiomics GTV (AUC, 0.88 [95% CI: 0.80, 0.96]; P = .01) models. It also achieved the best accuracy (93% [95% CI: 84%, 98%]). Both this model and the model with added perinodular features showed good calibration, whereas adding perinodular features did not improve the performance (integrated discrimination improvement, -0.02; P = .56). Conclusion Separating ground-glass and solid CT radiomic features of part-solid nodules was useful in diagnosing the invasiveness of lung adenocarcinoma, yielding a better predictive performance than the Brock, clinical-semantic, volumetric, and radiomics gross tumor volume models. Online supplemental material is available for this article. See also the editorial by Nishino in this issue. Published under a CC BY 4.0 license.

Predicting the invasiveness of lung adenocarcinomas appearing as ground-glass nodule on CT scan using multi-task learning and deep radiomics

Background

Due to different treatment method and prognosis of different subtypes of lung adenocarcinomas appearing as ground-glass nodules (GGNs) on computed tomography (CT) scan, it is important to classify invasive adenocarcinomas from non-invasive adenocarcinomas. The purpose of this paper is to build and evaluate the performance of deep learning networks on the differentiation the invasiveness of lung adenocarcinoma appearing as GGNs.

Methods

This retrospective study included 886 GGNs from 794 pathological confirmed patients with lung adenocarcinoma for training and testing the proposed networks. Three deep learning networks, namely XimaNet (deep learning-based classification model), XimaSharp (classification and nodule segmentation model), and Deep-RadNet (deep learning and radiomics combined classification model, i.e., deep radiomics) were built. Three classification tasks, namely task 1: classification of AAH/AIS and MIA, task 2: classification of MIA and IAC, and task 3: classification of non-invasive adenocarcinomas and invasive adenocarcinomas (AAH/AIS&MIA and IAC) were conducted to evaluate the model performance. The Z-test was used to compare the model performance.

Results

The AUC for classification of AAH/AIS with MIA were 0.891, 0.841 and 0.779 for Deep-RadNet, XimaNet and XimaSharp respectively. The AUC for classification of MIA with IAC were 0.889, 0.785 and 0.778 for three networks and AUC for classification of AAH/AIS&MIA with IAC were 0.941, 0.892 and 0.827 respectively. The performance of deep_RadNet was better than the other two models with the Z-test (P<0.05).

Conclusions

Deep-RadNet with the visual heat map could evaluate the invasiveness of GGNs accurately and intuitively, providing a theoretical basis for individualized and accurate medical treatment of patients with GGNs.

Prognostic value of epidermal growth factor receptor gene mutation in resected lung adenocarcinoma

BACKGROUND

Mutation of the EGFR gene is known as a predictor for the response to EGFR tyrosine kinase inhibitor. Although EGFR mutation status is proposed to be incorporated in the Ninth Edition of the Lung Cancer Staging system, its prognostic value for surgically resected lung adenocarcinoma remains controversial.

METHODS

Data on 1512 patients with completely resected lung adenocarcinoma who underwent EGFR mutation analysis between 2008 and 2015 were collected. The prognostic value of EGFR mutations was determined in patients with lung adenocarcinoma stratified by clinicopathologic and radiologic characteristics. Independent prognostic factors were identified by multivariate analysis using the Cox proportional hazards model. Competing risk model was used to estimate the cumulative incidence.

RESULTS

EGFR mutations were identified in 935 patients (61.8%). In the entire cohort, there was no difference in recurrence-free survival between the EGFR-mutated group and the wild-type group (P = .266). However, Cox multivariate analyses revealed that EGFR mutation was a strong independent prognostic factor for worse recurrence-free survival in patients with radiologic solid tumors (hazard ratio, 1.485; 95% confidence interval, 1.208-1.826; P < .001), histologic acinar pattern-predominant adenocarcinoma/papillary pattern-predominant adenocarcinoma/invasive mucinous adenocarcinoma (hazard ratio, 1.684; 95% confidence interval, 1.330-2.132; P < .001), and pathologic stage II and III (hazard ratio, 1.417; 95% confidence interval, 1.115-1.801; P = .004). Patients with EGFR mutations developed significantly more brain (hazard ratio, 1.827; 95% confidence interval, 1.213-2.766; P = .004) and bone (hazard ratio, 1.724; 95% confidence interval, 1.131-2.631; P = .011) metastases compared with the wild-type cohort.

CONCLUSIONS

EGFR mutation was a strong poor prognostic factor in patients with radiologic solid, histologic acinar pattern-predominant adenocarcinoma/papillary pattern-predominant adenocarcinoma/invasive mucinous adenocarcinoma, and pathologic stage II and III lung adenocarcinomas. After surgery, distinct metastatic patterns were revealed according to EGFR mutation status. These findings have implications for the upcoming new lung cancer staging system.

Management of Ground-Glass Opacities in the Lung Cancer Spectrum

BACKGROUND

Along with the popularity of low-dose computed tomography lung cancer screening, an increasing number of lung ground-glass opacity (GGO) lesions are detected. This review focuses on lung adenocarcinoma manifesting as GGO.

METHODS

We performed a literature search of the PubMed/MEDLINE database to identify articles reporting GGO. The following terms were used: GGO, ground-glass opacity, GGN, ground-glass nodule, part-solid nodule, and subsolid nodule.

RESULTS

GGO is a nonspecific radiologic finding showing a hazy opacity without blocking underlying pulmonary vessels or bronchial structures. The pathology of GGO can be benign, preinvasive, or invasive adenocarcinoma. Although radiographic features may indicate malignancy, a short period of follow-up is the optimal method to distinguish between benign and malignant GGO lesions. Pathologically, not only lepidic, but also nonlepidic growth patterns can present as GGO. Lung adenocarcinoma with a GGO component is associated with excellent survival compared with solid lesions. Moreover, there are distinct prognostic factors in patients with lung adenocarcinoma manifesting as GGO or solid lesions. For selected GGO-featured lung adenocarcinoma, sublobar resection with selective or no mediastinal lymph node dissection may be sufficient. Intraoperative frozen section is an effective method to guide resection strategy. A less intensive postoperative surveillance strategy may be more appropriate given the excellent survival. Management of multiple GGO lesions requires comprehensive considerations of GGO characteristics and patient conditions.

CONCLUSIONS

Lung adenocarcinoma manifesting as GGO defines a special clinical subtype with excellent prognosis. The management of GGO-featured lung adenocarcinoma should be distinct from that of solid lesions.

Computed tomography density is not associated with pathological tumor invasion for pure ground-glass nodules

OBJECTIVE

Pure ground-glass nodules are considered to be radiologically noninvasive in lung adenocarcinoma. However, some pure ground-glass nodules are found to be invasive adenocarcinoma pathologically. This study aims to identify the computed tomography parameters distinguishing invasive adenocarcinoma from adenocarcinoma in situ and minimally invasive adenocarcinoma.

METHODS

From May 2011 to December 2015, patients with completely resected adenocarcinoma appearing as pure ground-glass nodules were reviewed. To evaluate the association between computed tomography features and the invasiveness of pure ground-glass nodules, logistic regression analyses were conducted.

RESULTS

Among 432 enrolled patients, 118 (27.3%) were classified as adenocarcinoma in situ, 213 (49.3%) were classified as minimally invasive adenocarcinoma, 101 (23.4%) were classified as invasive adenocarcinoma. There was no postoperative recurrence for patients with pure ground-glass nodules. Logistic regression analyses demonstrated that computed tomography size was the only independent radiographic factor associated with adenocarcinoma in situ (odds ratio, 47.165; 95% confidence interval, 19.279-115.390; P < .001), whereas computed tomography density was not (odds ratio, 1.002; 95% confidence interval, 0.999-1.005; P = .127). Further analyses revealed that there was no distributional difference in computed tomography density among 3 groups (P = .173). Even after propensity score matching for adenocarcinoma in situ/minimally invasive adenocarcinoma and invasive adenocarcinoma, no significant difference in computed tomography density was observed (P = .741). The subanalyses for pure ground-glass nodules with 1 cm or more in size also indicated similar results.

CONCLUSIONS

In patients with pure ground-glass nodules, computed tomography size was the only radiographic parameter associated with tumor invasion. Measuring computed tomography density provided no advantage in differentiating invasive adenocarcinoma from adenocarcinoma in situ and minimally invasive adenocarcinoma.

A modified T categorization for part-solid lesions in Chinese patients with clinical stage I Non-small cell lung cancer

OBJECTIVES

We evaluated the prognostic impact of the presence of ground glass opacity (GGO) component and compared a modified clinical T categorization (cT_m) with the current 8th classification (cT₈) for survival prediction in Chinese patients with clinical stage I non-small cell lung cancer (NSCLC).

METHODS

According to cT_m and cT₈ classifications, we retrospectively evaluated 1461 patients with part-solid or pure-solid lesions. The recurrence-free survival (RFS) and overall survival (OS) were analyzed by Kaplan-Meier method and Cox proportional hazard model. The concordance index (C- index), reclassification improvement (NRI), integrated discrimination improvement (IDI), and decision curve analysis (DCA) were performed to estimate reclassification net benefits of cT_m for survival prediction.

RESULTS

The cT₈ classification clearly stratifies the survival outcomes in solid tumors but not in part-solid tumors. The presence of GGO components was an independent prognostic factor for both RFS and OS (p < 0.001), indicating a better outcome in each clinical T stage. The C-index was significantly improved from 0.650 to 0.730 for RFS (p < 0.001) and 0.647 to 0.730 for OS (p < 0.001) after reclassifying by cT_m categorization. The DCA, NRI (RFS: 0.342, OS: 0.302), and IDI (RFS: 0.070, OS: 0.054) demonstrated that the cT_m classification provided more net benefit in the survival prediction compared with the current cT₈ classification.

CONCLUSIONS

The current cT₈ classification may not be appropriate for part-solid lesions because the presence of GGO components is associated with excellent prognosis despite clinical stage. Also, the cT_m classification for part-solid lesions showed an improvement in survival prediction.

Prognostic impact of a ground-glass opacity component in clinical stage IA non-small cell lung cancer

OBJECTIVE

We performed a validation study to confirm the prognostic importance of the presence of a ground-glass opacity component based on data of the Japan Clinical Oncology Group study, JCOG0201, which was a prospective observational study to predict the pathological noninvasiveness of clinical stage IA lung cancer in Japan.

METHODS

Among the 811 patients registered in JCOG0201, 671 were confirmed eligible by study monitoring and a central review of computed tomography. Registered c-stage IA lung cancer was less than 30 mm in maximum tumor size, which was classified into a with ground-glass opacity group (pure ground-glass opacity and part-solid tumor) or solid group based on the status of a ground-glass opacity component. T staging was reassigned in accordance with the 8th edition of the TNM staging system. To validate the prognostic impact, overall survival was estimated.

RESULTS

Of the cases, 432 (64%) were in the with ground-glass opacity group and 239 (36%) were in the solid group with a median follow-up time of 10.1 years. The 5-year overall survival was significantly different between the with ground-glass opacity group and solid group (95.1% vs 81.1%). The 5-year overall survival was excellent regardless of the solid component size in the with ground-glass opacity group (c-T1a or less: 97.2%, c-T1b: 93.4%, c-T1c: 91.7%). In contrast, prognostic impact of the tumor size was definitive in the solid group (c-T1a: 87.5%, c-T1b: 85.9%, c-T1c: 73.7%).

CONCLUSIONS

Favorable prognostic impact of the presence of a ground-glass opacity component was demonstrated in JCOG0201. The presence or absence of a ground-glass opacity should be considered as an important parameter in the next clinical T classification.

Tumor Atelectasis Gives Rise to a Solid Appearance in Pulmonary Adenocarcinomas on High-Resolution Computed Tomography

Introduction

Ground-glass opacities in a high-resolution computed tomography (HR-CT) scan correlate, if malignant, with adenocarcinoma in situ. The solid appearance in the HR-CT is often considered indicative of an invasive component. This study aims to compare the radiologic features revealed in the HR-CT and the histologic features of primary adenocarcinomas in resection specimens to find the presence of tumor atelectasis in ground-glass nodules (GGNs) and part-solid and solid nodules.

Methods

HR-CT imaging was evaluated, and lung nodules were classified as GGNs, part-solid nodules, and solid nodules, whereas adenocarcinomas were classified according to WHO classification. Lepidic growth pattern with collapse was considered if there was reduction of air in the histologic section with maintained pulmonary architecture (without signs of pleural or vascular invasion).

Results

Radiologic and histologic features were compared in 47 lesions of 41 patients. The number of GGN, part-solid, and solid nodules were two, eight, and 37, respectively. Lepidic growth pattern with collapse was observed in both GGN, seven of the eight part-solid (88%) and 24 of the 37 solid (65%) lesions. Remarkably, more than 50% of the adenocarcinomas with a solid appearance in HR-CT imaging had a preexisting pulmonary architecture with adenocarcinoma with a predominant lepidic growth pattern. In these cases, the solid component can be explained by tumor-related collapse in vivo (tumor atelectasis on radiologic examination).

Conclusions

Tumor atelectasis is a frequent finding in pulmonary adenocarcinomas and may beside a ground glass opacity also result in a solid appearance in HR-CT imaging. A solid appearance on HR-CT cannot be attributed to invasion alone, as has been the assumption until now.

Validation of the Eighth Edition Clinical T Categorization System for Clinical Stage IA, Resected Lung Adenocarcinomas: Prognostic Implications of the Ground-Glass Opacity Component

INTRODUCTION

There is controversy regarding the clinical T (cT) category of lung adenocarcinomas that manifest as part-solid nodules (PSNs). We aimed to validate the cT category and to evaluate the independent prognostic role of the nodule type (i.e., part-solid versus solid).

METHODS

We retrospectively evaluated the prognostic value of clinico-radiologic factors regarding the overall survival of patients with clinical stage IA lung adenocarcinomas that were resected between 2008 and 2014. cT Category, nodule type, and their interaction term were included in the multivariable Cox regression analysis with other variables. In addition, a mixture cure model analysis was performed to investigate the association between the covariates and long-term survival.

RESULTS

A total of 744 patients (420 women; 362 PSNs; median age, 63 y) were included. The multivariable-adjusted hazard ratio (HR) of the nodule type was not significant (1.30, 95% confidence interval [CI]: 0.80-2.10, p = 0.291). However, the cT categories were significantly associated with overall survival (HR of cT1b, 2.33 [95% CI: 1.07-5.06, p = 0.033]; HR of cT1c, 5.74 [95% CI: 2.51-13.12, p < 0.001]). There were no interactions between the nodule type and the cT categories (all p > 0.05). The multivariable mixture cure model revealed that solid nodules were associated with a decreased probability of long-term survival (OR = 0.40, 95% CI: 0.18-0.92, p = 0.030). In addition, cT1c was a negative predictor of long-term survival (OR = 0.26, 95% CI: 0.07-0.94, p = 0.040).

CONCLUSIONS

The cT categorization system is valid for PSNs and solid nodules. Nevertheless, PSNs are a prognostic factor associated with long-term survival.

Imaging Features Suggestive of Multiple Primary Lung Adenocarcinomas

BACKGROUND

The tumor-node-metastasis classification system has proposed that lung cancers presenting as multifocal ground-glass nodules (multi-GGN) on computed tomography scan should be staged as multiple primaries instead of intrapulmonary metastases. However, the problem still exists for those synchronous multiple lung adenocarcinomas (SMLA) involving solid lesions. This study aimed to explore the distinct features of SMLA to better define the diagnosis and staging of this disease.

METHODS

Between 2008 and 2016, consecutive patients with complete resection of SMLA were prospectively enrolled in the study. The patients were divided into three groups based on CT images as follows: multi-GGN, one solid nodule plus one or more GGNs (solid-GGN), and multiple solid lesions with or without GGN (multi-solid). Clinicopathologic features and survival outcomes were compared between these groups. Multivariate Cox proportional hazards analyses using bootstrap internal validation were performed to identify independent predictors for recurrence-free survival (RFS) and overall survival (OS).

RESULTS

Of the 695 patients who met the inclusion criteria, 486 (69.9%) presented with multi-GGN tumor, 124 (17.9%) with solid-GGN tumor, and 85 (12.2%) with multi-solid tumor. The three groups had distinguished clinicopathologic features of gender, smoking history, nodal metastases, tumor size, subtype, and location (all P < 0.001). Multivariate analyses demonstrated that multi-solid tumor was an independent predictor for both decreased RFS [hazard ratio (HR) 2.941; 95% confidence interval (CI) 1.07-8.08; P = 0.036] and poor OS (HR 6.13; 95% CI 1.15-32.63; P = 0.034), but neither RFS (P = 0.384) nor OS (P = 0.811) differed between solid-GGN and multi-GGN tumors.

CONCLUSIONS

Both multi-GGN and solid-GGN tumors should be staged as multiple primaries, whereas multi-solid tumor was indicated to be advanced disease.

New Frontiers for Molecular Pathology

Lung cancer remains a serious oncological problem worldwide. The delayed diagnosis and a prevalence of advanced stages in up to 70% of cases at recognition are still observed. Thanks to targeted therapies and immunotherapy a significant progress in achieving prolonged survival in some lung cancer patients is reported. A precise histopathological diagnosis, especially the recognition of adenocarcinoma, and a progress in the methods of clinical staging underlie the proper qualification of patients for a tailored therapy. The deep molecular characteristics of lung cancer in liquid biopsy, for example blood, bronchoalveolar lavage fluid (BALF), cell suspension from needle aspirates, are currently available. The molecular characteristic has recently been extended with molecular aberrations of BRAF, KRAS, MET, ERBB2, RET, NTRK next to the well-known EGFR mutations and ALK, ROS-1 relocation. The present paper discusses the usefulness of adequate pathological methods and molecular testing for the identification of a broad spectrum of predictive biomarkers for a molecular-directed lung cancer therapy. Immunotherapy with immune checkpoint inhibitors (ICIs) is approved in the first line therapy of advanced non-small-cell lung cancer. To date only PD-L1 expression on tumor cells has been found to be a marker of response to ICIs. The efficacy of ICIs as well as the susceptibility to immune-related adverse events are highly individual, so immune biomarkers are widely investigated. The candidates for predictive factors for ICIs immunotherapy include cancer cell antigenicity, presence of regulatory/suppressory molecules on cancer cells, cancer stem cells or on exosomes, and, on the other hand, an immune status of the patient. Cancers with high immune infiltration in the tumor milieu, referred to as “hot” tumors, seem to ensure a better response to ICIs than the “cold” ones. BALF analysis may replace cancer tissue examination, which is of limited access in advanced stages, for the recognition of the nature of immune response in the tumor environment. Tumor mutational burden (TMB) was shown to correlate with a good response to ICIs, especially when combined with other anticancer therapies. The present paper demonstrates the results of recent studies on lung cancer characteristics which bring us closer to the definition of useful prognostic/predictive factors.

Role of PET/CT in Management of Early Lung Adenocarcinoma

OBJECTIVE. The purpose of this study was to explore the value of FDG PET combined with high-resolution CT (HRCT) in predicting the pathologic subtypes and growth patterns of early lung adenocarcinoma. MATERIALS AND METHODS. A retrospective analysis was conducted on the PET/CT data on ground-glass nodules (GGNs) resected from patients with stage IA lung adenocarcinoma. The efficacy of PET maximum standardized uptake value (SUV_max) combined with HRCT signs in prediction of histopathologic subtype and growth pattern of lung adeno-carcinoma was evaluated. RESULTS. SUV_max was significantly higher in GGNs with invasive HRCT signs. The diameter of GGN (odds ratio, 1.660; p = 0.000) and the difference in attenuation value (odds ratio, 1.012; p = 0.011) between ground-glass components and adjacent lung tissues were independent predictors of FDG uptake by GGNs. SUV_max was higher in invasive adenocarcinoma than in adenocarcinoma in situ (AIS)-minimally invasive adenocarcinoma (MIA) (median SUV_max, 2.0 vs 1.1; p = 0.008). An SUV_max of 2.0 was the optimal cutoff value for differentiating invasive adenocarcinoma from AIS-MIA. Acinar-papillary adenocarcinoma had a higher SUV_max than lepidic adenocarcinoma (median SUV_max, 2.1 vs 1.3; p = 0.037). An SUV_max of 1.4 was the optimal cutoff value for differentiating the growth pattern of adenocarcinoma. Use of PET/CT with HRCT significantly improved efficacy for differentiating invasive adeno-carcinoma from AIS-MIA. However, use of HRCT cannot significantly improve the diagnostic efficacy of FDG PET in the evaluation of tumors with different growth patterns. CONCLUSION. FDG PET can be used to predict the histopathologic subtypes and growth patterns of early lung adenocarcinoma. Combined with HRCT, it has value for predicting invasive histopathologic subtypes but no significance for predicting invasive growth patterns.

Distinct Prognostic Factors in Patients with Stage I Non-Small Cell Lung Cancer with Radiologic Part-Solid or Solid Lesions

INTRODUCTION

Recent studies have indicated that the presence of ground-glass opacity (GGO) components is associated with favorable survival. The purpose of this study was to reveal the prognostic value of GGO components and differences in prognostic factors for part-solid and solid lesions in invasive stage I NSCLC.

METHODS

The cases of 2010 patients with completely resected invasive pathological stage I NSCLC were reviewed according to the eighth edition of the TNM classification. Patients were categorized into the pure-GGO, part-solid, and solid groups based on consolidation-to-tumor ratio. Cox multivariate proportional hazard analyses were conducted to identify independent prognostic factors in each group.

RESULTS

Of the 2010 patients, 146 (7.3%) were in the pure-GGO group, 732 (36.4%) were in the part-solid group, and 1132 (56.3%) were in the solid group. Cox multivariate analyses revealed that GGO absence was a strong independent risk factor for worse recurrence-free survival (p < 0.001). For the pure-GGO group, there was no recurrence in spite of the invasive stage. For the part-solid group, visceral pleural invasion could not predict recurrence-free survival in general (p = 0.514) or in each tumor size group (for tumors size ≤1 cm, p = 0.664; for tumors size >1 to 2 cm, p = 0.456; for tumors size >2 to 3 cm, p = 0.900; and for tumors size >3 to 4 cm, p = 0.397). For the solid group, adenocarcinoma subtype was not a prognostic factor for recurrence-free survival in general (p = 0.162) or in each tumor size group (for tumors size ≤ 2 cm, p = 0.092; for tumors size >2 to 3 cm, p = 0.330; and for tumors size >3 to 4 cm, p = 0.885).

CONCLUSIONS

The presence of GGO components was a strong predictor in patients with invasive pathological stage I NSCLC. Risk factors were distinct in the part-solid and solid groups. There was no prognostic value of visceral pleural invasion in the part-solid group. Adenocarcinoma subtype did not have prognostic value in the solid group.