Gender and smoking differences in cell cycle marker expressions and p-TNM stage in non-small cell lung carcinoma

Authors


Correspondence
Kyung-Hee Kim, MD, Department of Pathology, Chungnam National University School of Medicine, 6 Munhwa 1-dong, Jung-gu, Daejeon 301-747, Korea. Email: phone330@cnu.ac.kr

ABSTRACT

Background and aim: A gender difference has been linked to the incidence and mortality of lung carcinomas. However, a comprehensive investigation including immunohistochemical studies of the gender difference involved in lung carcinoma progression has not been conducted. Methods: A total of 66 adenocarcinoma (AD) and 102 squamous cell carcinoma (SQ) samples were analyzed using immunohistochemistry for cell cycle-specific markers cyclin A, cyclin B1, cyclin D1 and Ki-67. Automated silver-enhanced in situ hybridization was used to evaluate epidermal growth factor receptor (EGFR) copy number. Results: For AD, male sex was significantly associated with the expression of cyclinA, cyclinB1 and high pathological tumor-node-metastasis (p-TNM) staging. For SQ, ever-smokers were associated with the expression of cyclin B1 and cyclin D1. For AD, ever-smokers were associated with the expression of cyclin A, cyclin B1 and Ki-67. There is no statistical significant correlation of smoking history with p-TNM stage and EGFR gene copy number in the AD or SQ, although the number of cases is limited. Conclusions: These results indicate that a gender difference contributes to AD growth and that smoking is associated with SQ and AD growth. The differential effects of gender and smoking differences may contribute through different pathways for AD and SQ subtypes.

INTRODUCTION

Lung carcinoma is the leading cause of cancer-related deaths in both men and women worldwide.1 Lung cancer has uniquely characteristic epidemiology and histology, for example, in the difference in incidence between men and women. Recently, some researchers have demonstrated the importance of gender difference in human lung cancers.2 Smoking women are more likely to develop lung adenocarcinoma (AD) than men3,4 even though more histological types of lung cancer occur in male smokers. The response to chemotherapy also differs with gender.5 These findings suggest that there might be gender differences in the molecular carcinogenic mechanism of lung cancer.

Dysregulation of the cell cycle is a distinct feature of human cancers.6 Progression of cell cycle is controlled by a series of cyclins and cyclin-dependent kinases (CDKs) and each cyclin acts at different phases of the cell cycle by binding and activating corresponding CDKs.7 Of the various cyclin/CDK complexes involved in cell cycle regulation, cyclin D1/CDK4/6 and cyclin B1/Cdc2 are of particular interest because the former directs G1-S-phase transition and the latter controls G2-M-phase checkpoint surveillance, which are essential events for DNA synthesis and cell proliferation.7 Cyclin A/CDK2 regulates events at the mitotic prophase. Dysregulated expression of these cyclins, CDKs, or both may lead to uncontrolled cell growth and malignant transformation. Amplification, overexpression, or both, of cyclin D1 has been reported in a large variety of human cancers, including those of the esophagus, head and neck, lung, liver, and breast,6 and is reported to be of prognostic importance in patients with most of these tumor types.8 Overexpression of cyclin B1 has been reported in breast, colon, prostate, oral, and esophageal carcinomas.9

Epidermal growth factor receptor (EGFR) is overexpressed in many human cancers, most notably non-small cell lung carcinoma (NSCLC).10 Hirsch et al.11 reported that increased EGFR gene copy number, assessed by fluorescence in situ hybridization, is associated with poor prognosis in NSCLC. Gefitinib, a novel molecular targeted drug, is an EGFR-tyrosine kinase inhibitor12 that was reported to be especially effective in treating Asian women with lung AD.13

Although many studies have been conducted, little is known about the roles of gender difference and smoking history in lung cancer and whether those factors may have potential clinical application. Thus, it is reasonable to propose that association of gender difference with the EGFR gene copy number status and cyclins, which, together with CDKs, regulate cell cycle progression, could lead to changes in cell proliferation. Therefore, the purpose of this study was to investigate the gender and smoking differences in cell cycle marker expression, EGFR gene copy number status, and pathologic tumor-node-metastasis (TNM) stage in NSCLC, especially in AD and squamous cell carcinoma (SQ).

METHODS

Patients, tissue samples, and reagents

We obtained 168 lung samples surgically resected between 1995 and 2008 from the surgical pathology files maintained at the Department of Pathology of Chungnam National University Hospital and Eulji University Hospital. The tissues consisted of 102 samples of SQ and 66 samples of AD. All samples were isolated from lobectomy or pneumonectomy specimens. Medical records were reviewed for all cases to obtain pertinent clinical information. This study was approved by the Institutional Review Board of Chungnam National University Hospital.

All cases were histologically reviewed by two pathologists (K.H.K. and Y.B.K.), and the two most representative areas of viable carcinoma tissue were selected and marked on the hematoxylin and eosin (H&E)-stained slides. To create a tissue microarray, tissue columns (3.0 mm in diameter) were punched from the original paraffin blocks and inserted into new recipient paraffin blocks (each containing 30 holes for tissue columns). Arrays were constructed using two 3-mm diameter cores per tumor.

Immunohistochemical staining

Tissue sections on microslides were deparaffinized with xylene, dehydrated using serial dilutions of alcohol, and heated in 10 mM sodium citrate (pH 6.0) using a pressure cooker at full power for 3 min for antigen retrieval. Peroxide blocking was performed using 3% H2O2 in methanol at ambient temperature for 10 min. To reduce background staining, sections were blocked in serum-free protein for 20 min. The sections were then incubated with the indicated primary antibodies at ambient temperature for 60 min. Primary antibodies were used at the following dilutions: rabbit polyclonal anti-cyclin A antibody (1:75; Santa Cruz Biotechnology, Santa Cruz, CA, USA), rabbit polyclonal anti-cyclin B1 antibody (1:75; Santa Cruz Biotechnology), rabbit monoclonal anti-cyclin D1 antibody (1:75; Thermo Scientific, Fremont, CA, USA), and mouse monoclonal anti-Ki-67 antibody (1:100; Dako, Glostrup, Denmark). After washing, samples were incubated in Dako REAL EnVision/HRP Rabbit/Mouse detection reagent (Dako) for an additional 20 min at room temperature, followed by additional washing. The chromogen was then developed for 2 min, and slides were counterstained with Meyer's hematoxylin, dehydrated, and mounted with Canada balsam.

Evaluation of immunostained samples

All immunostained slides were digitally scanned using a scanscope (Aperio ScanScope CS system, Vista, CA, USA). The central areas of each core, which contain the most preserved and representative tumor cells, were selected, as edge effects and staining artifacts can lead to data misinterpretation in peripheral areas.

For immunohistochemical staining for cyclin A, cyclin B1, cyclin D1 and Ki-67, nuclear or perinuclear staining was scored using digitally-scanned files and light microsopes. The proportion score (range, 0 to 5) and intensity score (range, 0 to 3) were added to obtain the total score (range, 0 to 8) (Fig. 1).14 Each sample was examined separately and scored by two pathologists (K.H.K. and Y.B.K.).

Figure 1.

Representative micrographs of cells showing epidermal growth factor receptor (EGFR) amplification, revealed by silver-enhanced in situ hybridization, cyclin A, cyclin B1, and cyclin D1 expressions, revealed by immunohistochemical staining. (a) Silver-enhanced in situ hybridization. Red signals correspond to the centromere of chromosome 7, and black signals correspond to the EGFR gene locus. (b) Strong nuclear expression of cyclin A. (c) Strong nuclear and perinuclear expression of cyclin B1. (d) Strong nuclear expression of cyclin D1.

Silver-enhanced in situ hybridization (SISH) to analyze EGFR amplification

For automated SISH, consecutive slides were stained according to the manufacturer's recommended protocol for the EGFR dual SISH-2p4 kit (Roche, Tucson, AZ, USA). This protocol was optimally formulated for use with XT Dual uView SISH RedISH v1 for the Benchmark® XT IHC/ISH Staining Module (Roche). The EGFR DNA probe (SISH V-probe) was denatured at 95°C for 12 min, and hybridization was performed at 52°C for 6 h. After hybridization, appropriate stringency washes (3 times at 72°C) were performed. The chromosome 7 probe (Red ISH V-probe) was denatured at 95°C for 12 min, and hybridization was performed at 44°C for 2 h. After hybridization, appropriate stringency washes (3 times at 59°C) were performed. The EGFR DNA probe was visualized using the rabbit anti-DNP primary antibody and ultraView SISH detection kit (Roche). The chromosome 7 probe was visualized using the rabbit anti-DNP primary antibody and the ultraView ISH AP detection Kit (Roche). Specimens were then counterstained with Ventana Hematoxylin II. Silver precipitation resulted in nuclear deposition of silver, and signals corresponding to the EGFR gene were visualized as black dots. The centromere of chromosome 7 was visualized as a red spot.

EGFR gene copy number was classified into six strata based on the frequency of tumor cells with increased EGFR gene copy numbers per cell (strata range, 1∼6): score 1, disomy (≤2 copies in >90% of cells); score 2, low trisomy (≤2 copies in ≥40% of cells, 3 copies in 10–40% of the cells, ≥4 copies in <10% of cells); score 3, high trisomy (≤2 copies in ≥40% of cells, 3 copies in ≥40% of cells, ≥4 copies in <10% of cells); score 4, low polysomy (≥4 copies in 10–40% of cells); score 5, high polysomy (≥4 copies in ≥40% of cells); and score 6, gene amplification (defined by presence of tight EGFR gene clusters and a ratio of EGFR gene to chromosome of ≥2 or ≥15 copies of EGFR per cell in ≥10% of analyzed cells). EGFR SISH positive was defined at score 5 or score 6 (Fig. 1).15,16

Statistical analysis

Statistical analyses for continuous variables were conducted using the Mann–Whitney U-test. Associations between categorical variables were assessed via cross-tabulation, the chi-square test and Fisher's exact test, and multivariable logistic regression analysis as indicated. The strengths of association between two variables were accessed via Spearman's coefficient of rank correlation. P-values less than 0.05 were regarded as statistically significant. All statistical analyses were performed using SPSS (version 17; SPSS, Chicago, IL, USA).

RESULTS

Clinicopathological features

Clinicopathologic features (age; smoking status; sex; expression of cyclin A, cyclin B1, cyclin D1, and Ki-67; EGFR copy number; and pathologic TNM staging) for AD and SQ samples are summarized in Table 1. In univariate analysis, statistically higher values of cyclin A, cyclin B1, and Ki-67 expressions were detected in the SQ patients when compared with the AD patients (P < 0.001, respectively). In Pearson's chi-square test, a higher proportion of ever-smoker was detected in the SQ than that of the AD patients (60.0% vs 30.4%, P < 0.001); a higher proportion of high pathologic TNM stage (III-IV) was detected in the AD patients than that of the SQ patients (39.4% vs 24.5%, P= 0.040) (Table 1). High Ki-67 expression was correlated with cyclin A and cyclin B1 for SQ (P= 0.005 and P < 0.001, respectively, Spearman's coefficient of rank correlation) and smoking status (never/ever smoker) was correlated with cycle B1 and cycle D1 for SQ (P= 0.014 and P= 0.040, respectively, Spearman's coefficient of rank correlation) (Table 2). For AD, high Ki-67 expression was correlated with smoking status (never/ever smoker), cyclin A and cyclin B1 (P= 0.032, P= 0.002, and P= 0.005, respectively, Spearman's coefficient of rank correlation) (Table 3).

Table 1.  Analysis comparing lung adenocarcinoma and squamous cell carcinoma with clinicopathological variables including expression of cell cycle markers (cyclin A, cyclin B1, cyclin D1 and Ki-67), and EGFR SISH results
CharacteristicsAdenocarcinomaSquamous cell carcinomaP-value
  1. EGFR, epidermal growth factor receptor; SISH, silver-enhanced in situ hybridization; No., total number of cases which were performed investigation; Mean ± SD, average score of the cases performed ± standard deviation; M, male; F, female; p-TNM, pathological tumor-node-metastasis.

  2. †Mann Whitney U-test; ‡Pearson's χ2 test.

Age  0.375†
 No.66102 
 Mean ± SD60.97 ± 10.43462.50 ± 9.706 
Smoking  <0.001‡
 No. (never/ever smoker)56 (39/17)90 (36/54) 
 Percentage100 (69.6/30.4)100 (40.0/60.0) 
Sex  <0.001‡
 No. (M/F)66 (38/28)102 (95/7) 
 Percentage100 (57.6/42.4)100 (93.1/6.9) 
EGFR SISH  0.260‡
 No.(negative/positive)47 (28/19)60 (42/18) 
 Percentage100 (59.6/40.4)100 (70.0/30.0) 
Cyclin A  <0.001†
 No.66102 
 Mean ± SD3.902 ± 1.5375.137 ± 1.199 
Cyclin B1  <0.001†
 No.66102 
 Mean ± SD3.720 ± 1.6185.206 ± 1.107 
Cyclin D1  0.666†
 No.66101 
 Mean ± SD3.636 ± 2.5193.856 ± 2.309 
Ki-67  <0.001†
 No.6596 
 Mean ± SD4.57 ± 2.2776.31 ± 1.468 
p-TNM stage  0.709‡
 No. (I/II-IV)66(33/33)102(54/48) 
 Percentage100 (50.0/50.0)100 (52.9/47.1) 
p-TNM stage  0.040‡
 No. (I-II/III-IV)66 (40/26)102 (77/25) 
 Percentage100 (60.6/39.4)100 (75.5/24.5) 
Table 2.  Correlation between cyclin A, cyclin B1, cyclin D1, and Ki 67 final scores by immunohistochemical staining, smoking status (ever/never smoker) and EGFR gene copy number status (negative/positive) by silver-enhanced in situ hybridization (SISH) of pulmonary squamous cell carcinomas
Spearman's rho Ki-67EGFR SISHCyclin ACyclin B1Cyclin D1
  1. EGFR, epidermal growth factor receptor.

  2. *P < 0.05; **P < 0.01.

EGFR SISHCorrelation coefficient−0.021    
Sig. (2-tailed)0.874    
No.59    
Cyclin ACorrelation coefficient0.282**−0.035   
Sig. (2-tailed)0.0050.790   
No.9660   
Cyclin B1Correlation coefficient0.386**−0.0700.719**  
Sig. (2-tailed)<0.0010.597<0.001  
No.9660102  
Cyclin D1Correlation coefficient0.0500.0140.219*0.189 
Sig. (2-tailed)0.6340.9170.0280.059 
No.9560101101 
SmokingCorrelation coefficient0.2020.0390.1870.259*0.218*
Sig. (2-tailed)0.0650.7930.0770.0140.040
No.8448909089
Table 3.  Correlation between cyclin A, cyclin B1, cyclin D1, and Ki-67 final scores by immunohistochemical staining, smoking status (ever/never smoker) and EGFR gene copy number status (negative/positive) by silver-enhanced in situ hybridization (SISH) of pulmonary of pulmonary adenocarcinomas
Spearman's rho Ki-67EGFR SISHCyclin ACyclin B1Cyclin D1
  1. EGFR, epidermal growth factor receptor.

  2. *P < 0.05; **P < 0.01.

EGFR SISHCorrelation coefficient−0.001    
Sig. (2-tailed)0.993    
No.47    
Cyclin ACorrelation coefficient0.377**−0.057   
Sig. (2-tailed)0.0020.704   
No.6547   
Cyclin B1Correlation coefficient0.343**−0.1110.708**  
Sig. (2-tailed)0.0050.457.000  
No.654766  
Cyclin D1Correlation coefficient−0.122−0.0450.060−0.112 
Sig. (2-tailed)0.3340.7630.6330.371 
No.65476666 
SmokingCorrelation coefficient0.289*0.0600.289*0.280*−0.021
Sig. (2-tailed)0.0320.7260.0310.0360.880
No.5537565656

Correlation of gender difference with expression of cell cycle-related proteins, EGFR gene copy number and p-TNM stage

For AD samples, male sex was significantly associated with the expression of cyclin A, cyclin B1 and p-TNM staging (I vs II-IV and I-II vs III-IV) (P= 0.017, P= 0.002, P= 0.003, and P= 0.040, respectively) (Table 4). Multivariate analyses, including smoking and age, showed that male sex appeared to be correlated with high p-TNM stage regardless age and smoking history (Table 5). There is no statistical significant association between gender difference and EGFR gene copy number status in AD (Table 4). For SQ, a study for correlation of gender difference with expression of cell cycle-related proteins and p-TNM stage was not available because of the limited numbers of female patients with SQ.

Table 4.  Association of gender difference in pulmonary adenocarcinoma with expression of cell cycle markers (cyclin A, cyclin B1, cyclin D1 and Ki-67), EGFR gene copy number status and pathologic TNM stage
CharacteristicsMaleFemaleP-value
  1. EGFR, epidermal growth factor receptor; TNM, tumor, node and metastasis; No., total number of cases which were performed investigation; Mean ± SD, average score of the cases performed ± standard deviation; SISH, silver-enhanced in situ hybridization; p-TNM, pathological tumor-node-metastasis.

  2. †Mann Whitney U-test; ‡Pearson's chi-square test.

Cyclin A  0.017†
 No.3828 
 Mean ± SD4.33 ± 1.2593.32 ± 1.706 
Cyclin B1  0.002†
 No.3828 
 Mean ± SD4.18 ± 1.4953.09 ± 1.587 
Cyclin D1  0.728†
 No.3828 
 Mean ± SD3.68 ± 2.5513.57 ± 2.519 
Ki 67  0.187†
 No.3827 
 Mean ± SD4.76 ± 2.3074.30 ± 2.250 
EGFR SISH status (%)28 (100.0)19 (100.0)0.847‡
 SISH-negative (%)17 (60.7)11(57.9%) 
 SISH-positive (%)11 (39.3%)8(42.1) 
p-TNM stage38 (100)28 (100)0.003‡
 I (%)13 (34.2)20 (71.4) 
 II-IV (%)25 (65.8)8 (28.6) 
p-TNM stage38 (100)28 (100)0.040‡
 I-II (%)19 (50.0)21 (75.0) 
 III-IV (%)19 (50.0)7 (25.0) 
Table 5.  Multivariable logistic regression analysis investigating the association of pathologic TNM stage with sex (male vs female), smoking (never-smokers vs ever-smokers) and age for lung adenocarcinoma
Variablep-TNM (stage I vs II-IV) Odds ratio (95% CI)P-valuep-TNM (stage I-II vs III-IV) Odds ratio (95% CI)P-value
  1. TNM, tumor, node and metastasis; CI, confidence interval.

Sex (male vs female)0.128 (0.028–0.586)0.0080.214 (0.052–0.887)0.034
Smoking0.278 (0.058–1.317)0.1070.450 (0.106–1.914)0.280
Age1.024 (0.969–1.081)0.4021.007 (0.955–1.061)0.810

Correlation of smoking status with expression of cell cycle-related proteins, EGFR gene copy number and p-TNM stage

For SQ patients, a smoking history (never-smokers vs ever-smokers) was significantly associated with the expression of cyclin B1 and cyclin D1 (P= 0.015 and P= 0.041, respectively) (Table 6). For AD patients, smoking history (never-smokers vs ever-smokers) was significantly associated with the expression of cyclin A, cyclin B1 and Ki-67 (P= 0.032, P= 0.038, and P= 0.034, respectively) (Table 7). There is no statistical significant correlation of smoking history (never-smokers vs ever-smokers) with p-TNM stage and EGFR gene copy number in the AD or SQ patients, although the number of cases is limited (Tables 6,7).

Table 6.  Association of smoking status in pulmonary squamous cell carcinoma with expression of cell cycle markers (cyclin A, cyclin B1, cyclin D1 and Ki-67), EGFR gene copy number status and pathologic TNM staging status
CharacteristicsNever-smokersEver-smokersP-value
  1. EGFR, epidermal growth factor receptor; TNM, tumor, node and metastasis; No., total number of cases which were performed investigation; Mean ± SD, average score of the cases performed ± standard deviation; SISH, silver-enhanced in situ hybridization.

  2. †Mann Whitney U-test; ‡Pearson's chi-square test; §Fisher's exact test.

Cyclin A  0.077†
 No.3654 
 Mean ± SD4.82 ± 1.4405.32 ± 1.092 
Cyclin B1  0.015†
 No.3654 
 Mean ± SD4.85 ± 1.3885.46 ± 0.905 
Cyclin D1  0.041†
 No.3653 
 Mean ± SD3.13 ± 2.4544.22 ± 2.246 
Ki 67  0.066†
 No.3648 
 Mean ± SD5.97 ± 1.8446.65 ± 0.785 
EGFR SISH status (%)35 (100.0)13 (100.0)1.000§
 SISH-negative (%)23 (65.7)8 (61.5) 
 SISH-positive (%)12 (34.3)5 (28.5) 
Pathologic stage36 (100)54 (100)0.120‡
 I (%)16 (44.4)33 (61.1) 
 II-IV (%)20 (55.6)21(38.9) 
Pathologic stage24 (100)42 (100)0.243‡
 I-II (%)12 (66.7)12 (77.8) 
 III-IV (%)15 (33.3)7 (22.2) 
Table 7.  Association of smoking status in pulmonary adenocarcinoma with expression of cell cycle markers (cyclin A, cyclin B1, cyclin D1 and Ki-67), EGFR gene copy number status and pathologic TNM stage
CharacteristicsNever-smokersEver-smokersP-value
  1. EGFR, epidermal growth factor receptor; TNM, tumor, node and metastasis; No., total number of cases which were performed investigation; Mean ± SD, average score of the cases performed ± standard deviation; SISH, silver-enhanced in situ hybridization.

  2. †Mann-Whitney U-test; ‡Pearson's χ2 test; §Fisher's exact test.

Cyclin A  0.032†
 No.3917 
 Mean ± SD3.72 ± 1.5594.71 ± 1.226 
Cyclin B1  0.038†
 No.3917 
 Mean ± SD3.59 ± 1.6344.32 ± 1.648 
Cyclin D1  0.878†
 No.3917 
 Mean ± SD3.77 ± 2.5023.56 ± 2.680 
Ki 67  0.034†
 No.3817 
 Mean ± SD4.00 ± 2.5575.41 ± 1.839 
EGFR SISH status (%)29 (100.0)8 (100.0)1.000§
 SISH-negative (%)17 (58.6)6 (75.0) 
 SISH-positive (%)12 (41.4)2 (25.0) 
Pathologic stage39 (100)17 (100)0.950‡
 I (%)21 (53.8)9 (52.9) 
 II-IV (%)18 (46.2)8 (47.1) 
Pathologic stage39 (100)17 (100)0.848‡
 I-II (%)24 (61.5)10 (58.8) 
 III-IV (%)15 (38.5)7 (41.2) 

DISCUSSION

In our study, three cyclins (cyclin A, cyclin B1 and cyclin D1), with known prognostic efficacy as cell cycle markers, and Ki-67, were used to investigate the relationships among potential risk factors: gender, histologic subtype (AD and SQ) and expression of EGFR gene copy number. In 1993, Risch et al.17 studied gender differences in the correlation of cigarette smoking with lung cancer. They found that compared with a nonsmoking population, the odds ratio of developing lung cancer for those with a 40 pack year history was 27.9 in women versus 9.6 in men.17 Other studies have demonstrated that the histological distribution differs between men and women.18 Kure et al.19 found a higher frequency of a specific mutation, G:C→T:A, in the p53 gene and a higher average DNA adduct level in the lung tumors of women and men, even though the level of exposure to carcinogens from cigarette smoking was lower among women (mean, 23 pack years) than among men (mean, 39 pack years).19 AD, especially pure bronchioloalveolar carcinoma characterized by a lepidic growth pattern, more indolent progression and lack of distant metastasis, comprised another distinct disease entity involving more females and more never-smokers.20,21

History of smoking was identified as a borderline statistically significant prognostic factor and shown to affect overall survival more significantly than sex in NSCLC.22 They found that patients with earlier stage, younger patients, never smokers and females had better overall survival in univariate analyses, but, in multivariate analyses, including sex, age and smoking, age appeared to be an independent significant prognostic factor with no statistical significance for sex and smoking. The median survival was the same between never-smoker women and men with AD with similar survival curves, but the data was not a statistical significance (P= 0.706).22 The differential effects of a gender and smoking differences in AD and SQ suggest that gender and smoking may contribute through different pathways in AD and SQ subtypes. Never-smokers generally included more females, were younger, with better performance status and had more AD diagnosed.23 In an exploratory analysis, different profiles were observed for predictive factors concerning chemotherapy response between the two groups. Never-smokers with NSCLC lived longer than ever-smokers. Gender, histology, and time of diagnosis are important factors for prognosis in these patients.23 As in previous studies,22,24 we found that women have a better prognostic factor (lower p-TNM stage). However, in the previous studies, survival advantages of female could be attributed to the younger age at the diagnosis and lower smoking prevalence. In our study, male sex with AD appeared to be an independent factor of high p-TNM stage according to the multivariate analyses, including sex, smoking and age (Tables 4,5). These results indicate that a gender difference is important to progression of pulmonary AD.

The association with smoking is stronger for pulmonary SQ.22 A previous study showed cyclin A expression in SQ was significantly higher than in AD.25 Consistent with these findings, the present study also showed that cyclin A expression and proportion of ever-smokers in SQ were higher than in AD (P= 0.000) (Table 1). A previous study showed that elevated levels of cyclin B1 expression might be an indicator of poor prognosis in NSCLC, particularly in AD.26 Our study showed that expression of Ki-67 was correlated with cyclin A and cyclin B1 expressions in AD. The overexpression of Ki-67 was correlated with cyclin A and cyclin B1 for SQ (P= 0.005 and P < 0.001, respectively) (Table 2). Some studies show, mean values of Ki-67 labeling index (LI) is higher for G2+G3 tumors than for G1 tumors. EGFR LI was higher for G1+G2 than for G3 tumors, and for pT3 than for pT1+pT2 tumors.27 Consistent with those findings, this study showed that high Ki-67 expression was correlated with EGFR gene copy number by SISH, cyclin A and cyclin B1 in SQ.

Certain limitation of the study should be addressed. The lack of female patients with SQ made it difficult to evaluate the association of gender difference in SQ with the expressions of cell-cycle markers, EGFR gene copy number status and pathologic TNM stage. The number of patients available in the evaluation of EGFR gene amplification was not enough to yield statistical validity. We did not include the studies of EGFR mutation and EGFR expression, continuous with the study of EGFR gene copy number.

In conclusion, statistically higher average scores of cyclin A and cyclin B1, and higher proportions of pathologic TNM stage were detected in the male patients with AD when compared with the female patients with AD. The ever-smokers accounted for higher expressions of cyclins, compared with the never smokers, for AD and SQ. These results indicate that there are gender differences in cell cycle marker expressions and p-TNM stage in AD and that smoking is associated with SQ and AD growth. The differential effects of gender and smoking differences may contribute through different pathways for AD and SQ subtypes. Therefore, the potential for differential prognosis based on sex and smoking differences may lead to new perspectives on therapeutic targeting of AD and SQ subtypes.

Ancillary