Analysis of mutational and clinicopathologic characteristics of lung adenocarcinoma with clear cell component

Introduction Lung adenocarcinoma with clear cell component is extremely rare and the cases reported in literature remain scarce. The biological behaviors, clinicopathologic characteristics, mutational status and prognosis of lung adenocarcinoma with clear cell component are still uncertain. Methods Thirty-eight lung adenocarcinomas with clear cell component and 1659 lung adenocarcinomas were subjected to the study. All the corresponding clinicopathologic data, the distributions of relapse-free survival (RFS) and overall survival (OS), and the status of gene mutations were investigated. Results Of 1697 adenocarcinomas, 38 (2.2%) had clear cell component. Fifty percent of adenocarcinomas with clear cell component (11/22) harbored EGFR mutation, 41 percent (9/22) harbored KRAS mutation and 5 percent (1/22) harbored AKT1 mutation. Univariable analysis revealed that sex, age, tumor stage, tumor size, nodal stage and pathology were all significant predictors of RFS and OS while the tumor size and nodal stage were still significant predictors in multivariable analysis. There were significantly differences in RFS and OS for lung adenocarcinomas with clear cell component compared with those lung adenocarcinomas. Conclusions Lung adenocarcinoma with clear cell component is a rare, malignant tumor with poor prognosis and displays more frequent EGFR and KRAS mutations.


INTRODUCTION
Lung cancer is the main cancer worldwide in terms of the incidence and mortality [1,2]. Accounting for nearly 50 percent of all lung cancers, adenocarcinoma remains the most common histologic subtype of lung cancer and the incidence rate continues to rise in virtually all countries [3,4]. In 2011, a new adenocarcinoma classification was proposed by the International Association for the Study of Lung Cancer (IASLC)/American Thoracic Society (ATS)/ European Respiratory Society (ERS), which discontinued the subtype of clear cell adenocarcinoma and recognized it as a feature [3].
Regardless of the amount and size, the term 'lung adenocarcinoma with clear cell component' is now regarded as a cytologic feature due to the IASLC/ ATS/ERS classification [3]. Adenocarcinoma cells with clear cytoplasm are frequently focal, but they rarely become the predominant component [5]. Many studies have demonstrated that clear cell feature may occur in association with various histologic subtypes even in some esoteric pulmonary tumors such as fetal adenocarcinoma [6][7][8]. Although clear cell feature is not considered to be a specific histologic pattern, associations with immunohistochemical and molecular characteristics were confirmed [3]. www.impactjournals.com/oncotarget Despite that clear cell feature may occur in multiple histologic subtypes of lung adenocarcinoma, to our knowledge, lung adenocarcinoma with clear cell component is extremely rare and the cases reported in literature remain scarce [9][10][11][12]. Furthermore, the biological behaviors, clinicopathologic characteristics, mutational status and prognosis of lung adenocarcinoma with clear cell component are still undetermined. Therefore, we undertook an investigation of lung adenocarcinoma with clear cell component and compared clinicopathologic characteristics between 38 lung adenocarcinomas with clear cell component and 1659 lung adenocarcinomas. We are the first to detect mutational status of lung adenocarcinoma with clear cell component in such a number.  Figure 1 and the characteristics of lung adenocarcinomas with clear cell component are listed in Table 1 and 2. Patients with lung adenocarcinomas consisted of 950 acinar-predominant (57%), 243 solidpredominant (15%), 312 papillary-predominant (19%), 80 lepidic-predominant (5%), 52 mucinous adenocarcinomapredominant (3%), 18 micropapillary-predominant (1%) and four not detected.

Clinicopathologic factors
Smoking history in lung adenocarcinomas with clear cell component and lung adenocarcinomas did not differ, but the former tended to have younger age (p <0.001), more male patients (p <0.001), larger tumor size (p <0.001), more advanced disease stage (p <0.001) and higher nodal stage (p <0.001) ( Table 1).

Survival analysis
Univariable analysis revealed that sex, age, tumor stage, tumor size, nodal stage and pathology were all significant predictors of RFS and OS ( Table 3). The tumor size and nodal stage were still significant predictors of RFS and OS in multivariable analysis, while pathology was not (Table 4).
During the follow-up, 17 (44.7%) lung adenocarcinomas with clear cell component and 525 (31.6%) lung adenocarcinomas experienced a relapse, and finally 13 (34.2%) and 294 (17.7%) patients died, respectively. There were significantly differences in RFS and OS for lung adenocarcinomas with clear cell component compared with those lung adenocarcinomas ( Figure 2).

DISCUSSION
Similar to the 2011 IASLC/ATS/ERS classification, a new classification of lung tumors was proposed by World Health Organization (WHO), which also defined clear cell feature as a type of cytologic characteristics [13]. Though many authors have outlined that clear cell feature can be detected in multiple histologic patterns and may be useful to compare diverse lung tumors, lung adenocarcinoma with clear cell component is extremely rare and the cases reported in literature remain scarce [6][7][8][9][10][11][12][14][15][16].
In our study, 38 lung adenocarcinomas with clear cell component were identified in1697 patients with primary lung adenocarcinoma (2.2%). The rarity is basically in line with prior studies reported by many authors [7,17,18].However, we found a slightly higher incidence rate than what Hinson et al. do (4 of 348, or 1.1%) [16]. Because they excluded the cases with clear cell component less than 50%, whereas we defined lung adenocarcinoma with clear cell component as tumor with     Figure 2). Nevertheless, this histologic pattern only acted as a significant predictor for survival in univariable analysis (Table 3) but not in multivariable analysis ( Table 4). The reason may be the fact that those lung adenocarcinoma with clear cell component are associated with advanced disease stage (stage III/IV, 39.5%, Table 1).
Overall  [7]. Although a previous study regarded lung adenocarcinoma with clear cell component as a tumor with a relatively good outcome, the study only had one patient and would produce a bias [9].
In consideration of the poor prognosis of lung adenocarcinoma with clear cell component, a thorough treatment strategy should be established. The treatment of clear cell tumor of the lung, a benign tumor, still maintains controversial [19][20][21]. As for lung adenocarcinoma with clear cell component, the main management for patients with early stage is surgical resection, while multidisciplinary treatment was needed in patients with advanced disease. In our study, we detected mutational status of lung adenocarcinomas with clear cell component. The findings revealed that 21 of 38 (55.3%) lung adenocarcinomas with clear cell component were detected harboring mutations in our tested genes, including 52% (11/21) EGFR mutations. And in another study, EGFR mutations and ALK rearrangements were also found in lung adenocarcinomas with clear cell component [6]. Whether EGFR tyrosine kinase inhibitors or ALK inhibitors are still effective for these special lung adenocarcinomas with clear cell component like lung adenocarcinoma is uncertain and calls for further investigation [13].
Despite clear cell feature is useful to compare diverse lung tumors, the feature can occur in a variety of patterns of lung tumor [14]. It is a challenging work to distinguish lung adenocarcinoma with clear cell component from multiple lung tumors with clear cell component, such as PEComa, benign (with clear cell tumor as a variant), squamous cell carcinoma with clear cell feature, primary clear cell carcinoma and metastases from renal clear cell carcinoma [9,13,22]. Differential diagnosis is necessary for these tumors mentioned above and histological and immunochemical tests play important roles. Tumor cells of lung adenocarcinoma with clear cell component may demonstrate finely vacuolated cytoplasm, reveal cytological and histological atypia recognized as lung adenocarcinoma [9,22]. PEComa exhibits strong cytoplasmic positivity for HMB-45 and is regarded as a benign tumor [23]. Squamous cell carcinoma with clear cell feature can be differentiated on histology and possesses typical immunohistochemical characteristics of squamous cell carcinoma positively expressing squamous markers p40 or p63 [13]. Primary clear cell carcinoma is usually Napsin A and TTF-1 negative [22,24,25]. Metastatic renal clear cell carcinoma is CD10 positive and the primary lesion can be found in the kidney by medical imaging technology [26,27].
There are several limitations in this study. First, despite we detected mutational status of lung adenocarcinomas with clear cell component and reported in such a number, it was still a small sample size. Besides, there were also a few lung adenocarcinomas with clear cell component not included in our study because of the incomplete data. Second, the controversial treatment strategies made us take different options to patients and would lead to diverse outcomes. This could be a potential bias.
In summary, our findings demonstrated that lung adenocarcinoma with clear cell component is a rare, malignant tumor with poor prognosis and displays more frequent EGFR and KRAS mutations. Multidisciplinary treatment and active postoperative follow-up should be adopted. Of the 1794 lung adenocarcinomas, 23 patients were excluded for receiving neoadjuvant chemotherapy. 74 patients were excluded because they were lost to follow-up. The remaining 1697 patients including 38 lung adenocarcinomas with clear cell component and 1659 lung adenocarcinomas were subjected to the study.

MATERIALS AND METHODS
An informed consent was signed by each patient or legal representative. The study started after obtaining our institutional review board approval. The archives for all patients were reviewed to collect corresponding clinicopathologic data including sex, age, smoking history, pathologic TNM stage (according to the 7th edition of the American Joint Committee on Cancer TNM staging system [28]) and treatment history. The data for disease recurrence and survival were obtained by follow-up clinic or telephone.

Statistical analysis
All the clinicopathologic data were analyzed using SPSS 19.0 software package (SPSS Inc, Chicago, IL) or Prism 5 (Graph Pad Software Inc., La Jolla, CA). The distributions of relapse-free survival (RFS) and overall survival (OS) were estimated using the Kaplan-Meier method, and the significance between two categories was probed by the log-rank test. The two-tailed significance level was set at p <0.05.