Differential expression of biomarkers in lung adenocarcinoma: a comparative study between smokers and never-smokers

T. Dutu1,2, S. Michiels3, P. Fouret4, F. Penault-Llorca5, P. Validire6, S. Benhamou7, E. Taranchon4,8, L. Morat2, D. Grunenwald9, T. Le Chevalier1, L. Sabatier2 and J.-C. Soria1,2,*

Departments of 1 Medicine, 8 Pathology, 3 Biostatistics and 7 CNRS UPR2169, 4 Translational Research Unit, Institut Gustave Roussy, Villejuif; 5 Department of Pathology, Centre Jean Perrin, Clermont-Ferrand; Departments of 9 Thoracique and 6 Pathology, Institut Mutualiste Montsouris, Paris; 2 Laboratoire de Radiobiologie et Oncologie, Commissariat à l'Energie Atomique, Fontenay aux Roses, France

* Correspondence to: Dr J.-C. Soria, Department of Medicine, Institut Gustave Roussy, 39 rue Camille Desmoulins, 94805 Villejuif, France. Tel: +33-1-42-11-43-01; Fax: +33-1-42-11-52-30; E-mail: soria{at}igr.fr


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Background: Non-small-cell lung cancer arising in never-smokers is usually of adenocarcinoma subtype. The oncogenic pathway of such tumors is poorly understood. To better define the biological characteristics of these tumors, we have compared the expression of a panel of epidermal growth factor receptor (EGFR)-related biomarkers in lung adenocarcinomas from smokers versus those in never-smokers.

Patients and methods: Using immunohistochemical analysis, we retrospectively analyzed EGFR, pAKT, PTEN, Ki-67, p27 and hTERT expression in specimens from 190 patients with completely resected lung adenocarcinomas (43 never-smokers and 147 smokers). These analyses were performed on tissue microarrays.

Results: EGFR expression was higher in tumors from smokers (P < 0.01), while pAKT was overexpressed mainly in tumors from never-smokers (P = 0.01). As expected, the tumors from smokers presented a higher expression of Ki-67 and a more frequent loss of expression of p27 (P < 0.01). In a multivariate model, two biological factors (p27 and Ki-67) and two clinical factors (age and sex) showed independent significant correlation with never-smoking status.

Conclusions: Lung adenocarcinomas in never-smokers have a very distinct immunohistochemical expression profile of EGFR-related biomarkers as compared with lung adenocarcinomas in smokers. High levels of EGFR and Ki-67 are observed in smokers, while never-smokers are characterized by high levels of pAKT and p27.

Key words: EGFR, immunohistochemistry, lung adenocarcinoma, never-smokers, pAKT


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Lung cancer is the most lethal neoplasia in the world and tobacco has been established as its main etiological agent [1Go]. Nevertheless, an increasing number of non-small-cell lung carcinomas (NSCLC) have been documented in never-smokers [2Go]. Environmental tobacco exposure, indoor and outdoor pollution, various carcinogens and genetic susceptibility have been incriminated as etiological factors for lung cancer of never-smokers [2Go]. While the carcinogenic process of smoking-related lung carcinomas is becoming well characterized, the pathogenesis of lung cancer in never-smokers is not. When compared with tobacco-related lung cancer, tumors from never-smokers are more frequently adenocarcinomas and occur with a disproportionately higher frequency in women [3Go]. Tumors from never-smokers often arise in a focal manner and are less frequently associated with gene mutations, loss of heterozygosity, chromosomal abnormalities and DNA methylation than tobacco-related lung tumors [4Go–7Go]. Despite these numerous differences that could sign specific genetic alterations, traditionally, the treatment of NSCLC does not take generally into account the tobacco status of patients. One major change might, however, arise from the use of epidermal growth factor receptor tyrosine-kinase inhibitors (EGFR TKI). Indeed, EGFR TKI have higher response rates in NSCLC from never-smokers, notably in the adenocarcinoma subtype [8Go, 9Go]. The present work aims to better characterize the EGFR-related signaling pathways in resected adenocarcinomas from never-smokers as compared with those from smokers. A panel of candidate biomarkers was chosen to be analyzed in this clinical population by means of immunohistochemistry.

EGFR mediates cancer cell growth, proliferation, angiogenesis, invasion and metastasis, and inhibits apoptosis [10Go]. In NSCLC, overexpression of EGFR has been described in 90% of squamous cell carcinomas and in 30–65% of adenocarcinomas [11Go], and has been variably correlated with clinical outcome [11Go, 12Go]. Some authors have stated that EGFR expression is not associated with EGFR TKI sensitivity [13Go, 14Go]. However, recently, two important studies have suggested that EGFR expression evaluated by immunohistochemistry can be associated with gefitinib [15Go] and erlotinib [16Go] sensitivity. On the other hand, the response to EGFR TKI (which is noted in up to 20% of treated patients) appears to be more frequent in some clinical groups like Japanese patients [17Go], adenocarcinoma subtypes, women and never-smokers [8Go, 9Go]. High EGFR gene copy number identified by fluorescence in situ hybridization may also be an effective molecular predictor for gefitinib efficacy in advanced NSCLC [15Go].

Phosphatidylinositol 3-kinase (PI3K)/AKT is a major signaling pathway that mediates EGFR effects on proliferation and survival. AKT is constitutively active in the majority of NSCLC cell lines [18Go] and high levels of phosphorylated (activated) AKT have been found in the majority of patients with lung carcinoma [19Go]. PTEN is a tumor suppressor gene that physiologically down-regulates the AKT pathway [20Go]. PTEN inactivation is found in a great number of cancers and in up to 25% of NSCLC [20Go, 21Go]. Ki-67 is the most common biomarker used to characterize tumoral proliferation. In NSCLC expression of Ki-67 frequently correlates with poor prognosis [22Go, 23Go]. p27 plays a crucial role in the regulation of the cellular cycle by inhibiting cdk2–cyclin E and thus controlling G1–S transition [24Go]. Additionally, p27 is a putative tumor-suppressor gene, regulator of drug resistance in solid tumors and promoter of apoptosis [24Go]. Loss of expression of p27 has been described in up to 58% of NSCLC [24Go–26Go], especially in smokers [26Go], and may correlate with a poor prognosis. Telomerase is a ribonucleoprotein responsible for telomere stabilization and cellular immortality; nucleolar immunohistochemical staining of the catalytic subunit (hTERT) of telomerase is largely restricted to cells with telomerase activity [27Go, 28Go]. hTERT expression may be stimulated by tobacco exposure, even in the absence of invasive carcinoma [29Go, 30Go]. Furthermore, telomerase expression is partially under the control of EGFR, while itself modulates epidermal growth factor base level [31Go, 32Go].

The implication of these biomarkers in NSCLC has been largely described, but limited data exist as for their differential expression in resected lung adenocarcinoma from smokers versus never-smokers.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Patients
Two hundred and twenty-nine consecutive patients who had undergone curative surgical treatment for lung adenocarcinoma at Institut Mutualiste Montsouris between 1995 and 2003 were identified. Patients who had received either neo-adjuvant chemotherapy or radiation therapy before surgery were excluded (n = 32).

Patients' records were retrospectively reviewed for smoking exposure as assessed by trained physicians at diagnosis. A patient was considered as a smoker if he admitted any active smoking history. Thus the ‘smoker’ category included ex-smokers as well as current smokers. Patients who denied any active smoking exposure were defined as never-smokers. Seven patients were excluded because their smoking history was unavailable. Finally, our study population consisted of 190 patients (147 smokers and 43 never-smokers). The patients' records and biomarker analysis were carried out with ethics committee approval.

Tissue microarrays
A central review by a pathologist (P.V.) confirmed the diagnosis of adenocarcinoma in all hematoxylin–eosin (HE)-stained sections. One paraffin-embedded adenocarcinoma sample was randomly selected for each case. Three representative tumor areas without necrosis were carefully selected on the corresponding HE-stained slide, and were subsequently punched for tissue microarrays. Blocks were formalin-fixed and paraffin-embedded.

Tissue microarrays production was performed by the Department of Pathology at Institut Gustave Roussy. All 190 tumors were punched in triplicate and cores 0.6 mm in diameter and 5 mm in length were inserted into a recipient paraffin block every 1 mm. Consecutive 5-µm sections were cut from the arrays and mounted onto charged slides. Every five sections, a slide was stained by HE to confirm the presence of the histological features of interest.

Immunohistochemistry
Sections of TMA blocks were stained using monoclonal or polyclonal antibodies against EGFR, pAKT, Ki-67, p27, hTERT and PTEN (Table 1). Deparaffinized sections were either heated in 10 mM citrate buffer (pH 6.0) or partially digested with proteinase K (Table 1). Sections were then immersed in methanol containing 0.3% hydrogen peroxide for 15 min to block the endogenous peroxidase activity and were incubated during 30 min thereafter with 5% blocking horse serum to reduce non-specific binding. Sections were incubated with the primary antibody according to previously validated conditions (Table 1). Following several washes with phosphate-buffered saline, the slides were processed for 30 min with a universal secondary biotinylated antibody, then with avidin-biotinylated horseradish peroxidase H complex (ABC kit; Vector Laboratories, Burlingame, CA, USA). Diaminobenzidine was used as a chromogen (8–12 min), and hematoxylin was used for counterstaining. Staining was performed simultaneously on all TMA slides. For each reaction, sections of lung cancer previously validated by us to be strongly positive for the assayed antibody were included as positive controls. As negative control, the staining procedure was performed with the primary antibody omitted. For PTEN staining, endothelial cells were also used as an internal control.


View this table:
[in this window]
[in a new window]
 
Table 1. Immunohistochemical conditions

 
Scoring of stained slides
Analysis of stained sections was performed using Spot Browser software (ALPHELYS, Plaisir, France). A minimum of 500 tumor cells were counted. If differences occurred between spots intensity, the most positive spot was taken into account. Subcellular distribution was used to define positivity as follows: membranous (EGFR), cytoplasmic (PTEN), nuclear (pAKT, Ki-67, p27) and nucleolar (hTERT). Heterogenous staining was noted for EGFR, pAKT, Ki-67, hTERT and p27. Therefore, in order to define positive cells, we first used a qualitative score (absent, low, moderate or strong staining) and then we assessed the percentage of positive cells (i.e. with moderate or strong staining) for each tumor, as described previously [23Go, 26Go, 33Go–35Go]. Thus, we will herein refer to these five biomarkers as ‘semi-quantitative’. Since PTEN staining was either homogenously positive or negative, we used a qualitative score (0 = absent, 1 = present PTEN staining).

Statistical analysis
Statistical analyses were performed with the SAS system version 8.2 [36Go]. Data clusters were visualized using Euclidian distance and average linking computed with GENESIS version 1.2.1 [37Go].

We used the Fisher's exact, {chi}2 and Wilcoxon two-sample tests to assess differences between the clinicopathological variables of patients who had a smoking history or not. For each semi-quantitative biomarker, univariate analysis of the immunohistochemical expression according to smoking status and other clinical variables was performed using the Wilcoxon two-sample test, comparing the location (median) of the expression values.

Logistic regression models were applied to compare the expression of biomarkers according to smoking history while adjusting for the other clinical variables. For the logistic regression analysis, the expression of each of the five semi-quantitative biomarkers was dichotomized into two classes: expression value below or equal to the median value of the series of 190 patients and expression value above the median value of the series. A multivariate logistic regression model was constructed applying a backwards stepwise selection procedure. Relationships between the expressions of the biomarkers were studied using Spearman's rank order correlation coefficient and tested by Spearman's rank test. The significance level chosen was P < 0.05 and all tests were two-sided.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Patient's characteristics
A total of 190 patients with resected lung adenocarcinoma were included in the present analysis: 147 patients were current or former smokers and 43 patients were never-smokers. The characteristics of these patients are described and compared in Table 2. In our series, never-smokers diagnosed with adenocarcinoma were more frequently female and were older than smokers. Tumors from never-smokers were more frequently well-differentiated. Less differentiated tumors were also found in patients with advanced stages (TNM > I): 61% versus 41% (P < 0.01). There was a significant relationship between tumor differentiation and gender: 63% of women (57 cases) versus 47% of men (47 cases) had well-differentiated tumors (P = 0.02).


View this table:
[in this window]
[in a new window]
 
Table 2. Patients' characteristics

 
Immunohistochemical staining
The characteristic patterns of immunohistochemical staining for EGFR, pAKT, Ki-67, p27, hTERT and PTEN are illustrated in Figure 1. Subcellular staining was noted predominantly in membranes (EGFR), cytoplasms (PTEN), nuclei (pAKT, Ki-67, p27) and nucleoli (hTERT) of tumors cells.



View larger version (138K):
[in this window]
[in a new window]
 
Figure 1. Representative immunohistochemical stainings for epidermal growth factor receptor (EGFR) (A), pAKT (B), Ki-67 (C), p27 (D), hTERT (E) and PTEN (F). The pattern was membranous (EGFR), cytoplasmic (PTEN), nuclear (pAKT, Ki-67, p27) and nucleolar (hTERT).

 
Relationship between individual markers and smoking status
For each semi-quantitative biomarker, the immunohistochemical expression was dichotomized into two classes: expression lower or equal than the median value or expression higher than the median value across the 190-patient series (the median expression values were 6% for Ki-67, 32% for p27, 50% for EGFR, 46% for hTERT and 35% for pAKT). PTEN expression was kept in the two original ordered classes. The comparison of immunohistochemical staining for the semi-quantitative biomarkers between smokers and never-smokers is presented in Figure 2 and Table 3.



View larger version (17K):
[in this window]
[in a new window]
 
Figure 2. Unsupervised hierarchical clustering of the 190 tumors based on the similarities in expression measured over the semi-quantitative biomarkers. Two-dimensional representation of immunohistochemical expression for 190 tumors obtained by hierarchical clustering for both biomarkers as for tumors (complete linkage). For each biomarker immunohistochemical expression was first normalized by subtracting the median value and dividing by the standard deviation across the 190 patients. The distance metric used was 1 minus Spearman's rank correlation. Each column represents a single biomarker and each row a tumor. As shown in the color bar, red indicates a high level of expression in the tumor, as compared with the median value; green indicates a low level of expression; and black no change.

 

View this table:
[in this window]
[in a new window]
 
Table 3. Immunohistochemical results in smokers versus never-smokers

 
EGFR expression was higher in tumors from smokers (P < 0.01), and pAKT was overexpressed mainly in tumors from never-smokers (P = 0.01). Tumors from smokers presented a higher expression of Ki-67 and a more frequent loss of expression of p27 (P < 0.01). PTEN and hTERT expression was not significantly different between smokers and never-smokers when adjusting for age, gender and differentiation.

Among all smokers the median number of pack-years was 40 (range 5–120); this value was 40 for men and 35 for female (Wilcoxon test P = 0.06). When testing the correlation between immunohistochemical staining and the degree of smoking as measured by pack-years, only Ki-67 expression was significantly correlated and this correlation was positive (Spearman's rank test P < 0.01).

Relationship between markers and other clinical variables
We investigated the univariate relationship between immunohistochemical expression and clinical variables such as gender, age, TNM stage and tumor differentiation (Table 4). The expression of EGFR, Ki-67 and hTERT was significantly higher in males than in females (P ≤ 0.03). Ki-67 was significantly higher expressed in tumors with advanced disease (TNM stage II–IIIa), while p27 was significantly underexpressed (P < 0.01). pAKT and hTERT (P < 0.01) were significantly underexpressed in those patients as compared with patients with limited disease (TNM stage I).


View this table:
[in this window]
[in a new window]
 
Table 4. Univariate analysis between the biomarkers' immunohistochemical expression and clinical characteristics

 
The expression of Ki-67, pAKT and p27 was significantly associated with tumor differentiation. Only EGFR and Ki-67 were significantly correlated with age, elderly patients expressing less EGFR and Ki-67 than the younger ones.

Unsupervised analysis
A global representation of the immunohistochemical profile of all analyzed biomarkers in the 190 patients studied is displayed in Figure 3. Unsupervised hierarchical clustering of the study series based on the similarities in expression measured over the five semi-quantitative biomarkers revealed two distinct main clusters of patients (vertical axis) with a significantly different proportion of non-smokers (P < 0.001), i.e. a top cluster consisting of only a few non-smokers (seven out of 96) and a bottom cluster with more non-smokers (36 out of 94). The bottom cluster of patients were characterized by a low expression of EGFR and Ki-67 and a high expression for p27, pAKT and hTERT. When taking into account biomarker clusters (horizontal axis), Ki-67 and EGFR were clustered together, as well as p27 and pAKT.



View larger version (16K):
[in this window]
[in a new window]
 
Figure 3. Box plots for each of the semi-quantitative biomarkers according to the smoking history. The grey box shows the limits of the middle half of the data (the white line inside the box represents the median expression value). Whiskers are drawn to the nearest value not beyond a standard span (1.5 interquantile range) from the quartiles. Extreme points are highlighted using crosses. P values are obtained with logistic regression models modeling the probability of never-smokers while adjusting for age, sex and differentiation.

 
Multivariate analysis
Before proceeding with the multivariate analysis, we explored the correlation between the expression of the semi-quantitative biomarkers (Table 5). Applying Spearman's rank test, we found a significant positive correlation (P ≤ 0.05) between the expression of EGFR and Ki-67, between Ki-67 and hTERT, between p27 and hTERT, and between pAKT and both p27 and hTERT.


View this table:
[in this window]
[in a new window]
 
Table 5. Spearman correlation values between the five semi-quantitative biomarkers (P values obtained with Spearman's rank test)

 
We determined a multivariate logistic regression model including clinical variables as well as biomarkers. A backwards stepwise procedure was applied starting from the model with all clinical and biological variables. We included as clinical parameters the tobacco status, age, gender, TNM stage and tumor differentiation. The expression of EGFR, pAKT, p27, Ki-67, hTERT and PTEN represented the biological variables. Variables were excluded from the model if P > 0.05. The logistic regression analysis was built using never-smoking status as the event. As presented in Table 6, two biological factors (p27 and Ki-67) and two clinical factors (age and gender) showed independent significant correlation with never-smoking status. Overall, the probability of an association with the never-smoking status was 4.5-fold higher for tumours with a high p27 score (i.e. above the median) than those with a low score. Conversely, the odds ratio (OR) for Ki-67 was 0.1, indicating that cases with high Ki-67 score had almost 10-fold less chances to be never-smokers than those with low Ki-67 scores. As for age, the probability for an older patient to be a never-smoker was 2.1-fold higher than the one in a patient 10 years younger. As expected in our study population, women had a disproportionate higher probability of being non-smoker than men (OR 91.9).


View this table:
[in this window]
[in a new window]
 
Table 6. OR estimates for the probability of never-smokers as compared with smokers in the multivariate model obtained by the stepwise selection procedure

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
This is the largest series comparing the immunohistochemical expression of EGFR, pAKT, Ki-67, p27, hTERT and PTEN in lung adenocarcinomas from smokers and never-smokers. Our results indicate that the pattern of expression of all biomarkers but PTEN is different between smokers and never-smokers, confirming that specific abnormalities, including specific changes in EGFR signaling pathways characterize lung carcinogenesis in never smokers.

We found an enhanced expression of EGFR in tumors from smokers as compared with those from never-smokers. This confirm previous observations suggesting that EGFR expression is enhanced by tobacco exposure [12Go]. It is largely accepted that the immunohistochemical expression of EGFR in NSCLC does not predict the response to pharmacological agents inhibiting this receptor [13Go, 14Go]. In line with these findings, high levels of EGFR expression have been reported in squamous cell carcinomas, while response to EGFR inhibitors is higher in lung adenocarcinomas [12Go]. Nevertheless, recently, two important studies demonstrated that EGFR expression evaluated by immunohistochemistry associated with gefitinib [15Go] and erlotinib [16Go] sensitivity. However, more than the degree of EGFR overexpression, the effectiveness of EGFR inhibitors may correlate with the tumor dependency on the EGFR pathway. In this regard, presence of activating mutations of EGFR tyrosine kinase have been reported to correlate with adenocarcinoma subtype [38Go], never-smoker status [9Go, 38Go], and a higher sensitivity to gefitinib or erlotinib [39Go–41Go].

Activation of EGFR subsequently triggers several signal transduction cascades including the PI3K-AKT pathway. Activation of AKT can be induced by tobacco exposure [42Go] and may be frequent in preneoplasic bronchial lesions in smokers [43Go]. A positive correlation has been reported between the expression of EGFR and activated downstream pAKT and pERK expression in NSCLC [19Go]. Our results suggest that pAKT might be involved in the development of lung adenocarcinoma in never-smokers, as tumors from such patients have a higher expression of pAKT than those from smokers. In large series of patients with NSCLC treated by gefitinib, other authors have also indicated that strong pAKT expression might correlate with non-smoking status and with an enhanced response to EGFR inhibitors [33Go]. Thus, it is possible that the development of lung cancer in never-smokers depends more on the EGFR–pAKT pathway than in smokers, in whom the elevated numbers of genomic and molecular abnormalities are responsible for a more complex carcinogenesis. However, in our study, pAKT positivity is not associated with EGFR expression either in smokers or in never-smokers.

PTEN is a negative regulator of the PI3K/AKT pathway [44Go]. Genomic alterations and loss of expression of the PTEN protein have been found in several malignancies including NSCLC [20Go, 21Go] but no data exist about PTEN expression according to smoking exposure. Similar to published reports, we found lack of PTEN expression in 10% of analyzed tumors, which is concordant with the frequency previously reported in NSCLC [21Go, 45Go]. In our study, smoking status did not correlate with the lack of PTEN expression. This suggests that alternative mechanisms (i.e. other than PTEN inactivation) may be responsible for the higher expression of pAKT in tumors from never-smokers.

Our results also suggest that lung adenocarcinoma of smokers tend to be more proliferative than those of never-smokers, as higher levels of expression of Ki-67 and lowers levels of expression of p27 were observed in tumors occurring in patients with a history of active tobacco exposure. The relationship between Ki-67 expression in lung cancers and the cigarette smoking status of patients has been previously reported [23Go, 46Go]. Higher expression levels of Ki-67 correlated with the degree of smoking (measured by the number of pack-years) in our study that included a large number of patients with adenocarcinoma. Several authors have reported that the immunohistochemical expression of p27 may represent a prognostic factor in NSCLC and could be correlated with the tobacco exposure history [25Go, 26Go]. Compared with our series, these studies had a limited number of patients with adenocarcinomas along with lower frequencies of loss of p27 expression. Our study confirms in a histology-homogenous population that p27 expression is lower in smokers than in never-smokers using a logistic regression analysis adjusted for age, gender and tumor differentiation. Conversely, we have not found any correlation between pack-years and p27 expression nor between the loss of p27 and the expression of Ki-67. The lack of correlation between these two biomarkers may be related to the fact that numerous other abnormalities beyond p27 expression (cyclins, cdk, p53, Rb) do also contribute to tumor proliferation, overall estimated by Ki-67 levels.

Telomerase and its catalytic component hTERT are expressed in the vast majority of NSCLC [35Go, 47Go, 48Go]. It has been postulated that tobacco exposure may be responsible for a great number of DNA adducts in the guanine-rich regions [49Go], and thus preferentially in the telomeric DNA, which is made by TTAGGG repetitive sequence [29Go]. The tobacco-induced alterations of telomeres may be responsible for a frequent activation of telomerase in smokers [29Go]. In line with published reports, our study demonstrates that hTERT expression in lung adenocarcinomas is higher in smokers than in never-smokers [29Go, 30Go, 50Go]. However, this difference is mainly related to the higher expression of hTERT among males as compared with females.

Our retrospective analysis raises several questions. First, a potential cause of bias in our analysis relates to the characterization of patients as smokers and never-smokers. Traditionally, never-smokers are defined as individuals having a lifetime exposure of fewer than 100 cigarettes. We consistently assessed the number of pack-years for each patient before surgery and were able to accurately discriminate between smoker and never-smoker individuals. The frequency of never-smokers in our population of patients with adenocarcinomas (23%) is comparable to the range found in other published studies [6Go]. Secondly, we noted an imbalance between the frequencies of never-smokers in male as compared with female patients. Even if lung adenocarcinoma occurs rarely in male non-smokers subjects, in our study, we decided not to limit ourselves to women with adenocarcinoma. We selected all consecutive intention-to-treat cases from a single institution operated for lung cancer with a final diagnosis of adenocarcinoma. Even though we expected a higher frequency of never-smokers in females, we were nonetheless surprised that in our series only 2% of men were never-smokers compared with 45% in women. However, such an imbalance has been reported previously in patients with lung cancer [6Go], and adenocarcinoma subtype in particular. One potential hypothesis for this imbalance is that adenocarcinomas in never-smoker males might be associated with a more advanced stage, thus excluding them from surgical series. Thirdly, we used commercially available antibodies validated for clinical or research applications. We reported median percentages of stained cells in order to discriminate between the different groups. This has allowed us to present a more comprehensive and hierarchical clustering for both biomarkers and tumors using a two-dimensional representation. While our data are not directly linked to any genomic analysis, we believe that clustering of immunohistochemical results according to EGFR mutational status is warranted. In fact, our group is undertaking a collaborative mutational analysis of HER1, HER2, PI3K and RAS in this very same patient population. This should help to elucidate the role of tobacco-exposure in inducing specific gene mutations such as the ones recently described in lung adenocarcinoma.

The development of NSCLC in never-smokers remains largely unknown. Several etiological [2Go], genetic [4Go–7Go] and molecular [19Go, 23Go, 33Go, 39Go] differences have been reported between NSCLC in never-smokers and tobacco-related NSCLC. Our results contribute to improve the description of lung carcinogenesis and highlight that etiology may be important for identifying better prognostic and predictive factors in NSCLC.


    Acknowledgements
 
The work in L.S.'s laboratory was supported by EDF and RISC-RAD contract number FI6R-CT2003-508842. T.D. was the recipient of a DUERCC grant.

Received for publication July 1, 2005. Revision received July 28, 2005. Accepted for publication August 2, 2005.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
1. Parkin DM, Bray F, Ferlay J et al. Estimating the world cancer burden: Globocan 2000. Int J Cancer 2001; 94: 153–156.[CrossRef][ISI][Medline]

2. Alberg AJ, Samet JM. Epidemiology of lung cancer. Chest 2003; 123: 21S–49S.[CrossRef][ISI][Medline]

3. Kabat GC. Aspects of the epidemiology of lung cancer in smokers and nonsmokers in the United States. Lung Cancer 1996; 15: 1–20.[CrossRef][ISI][Medline]

4. Powell CA, Bueno R, Borczuk AC et al. Patterns of allelic loss differ in lung adenocarcinomas of smokers and nonsmokers. Lung Cancer 2003; 39: 23–29.[CrossRef][ISI][Medline]

5. Sanchez-Cespedes M, Ahrendt SA, Piantadosi S et al. Chromosomal alterations in lung adenocarcinoma from smokers and nonsmokers. Cancer Res 2001; 61: 1309–1313.[Abstract/Free Full Text]

6. Toyooka S, Maruyama R, Toyooka KO et al. Smoke exposure, histologic type and geography-related differences in the methylation profiles of non-small cell lung cancer. Int J Cancer 2003; 103: 153–160.[CrossRef][ISI][Medline]

7. Wong MP, Lam WK, Wang E et al. Primary adenocarcinomas of the lung in nonsmokers show a distinct pattern of allelic imbalance. Cancer Res 2002; 62: 4464–4468.[Abstract/Free Full Text]

8. Kris MG, Natale RB, Herbst RS et al. Efficacy of gefitinib, an inhibitor of the epidermal growth factor receptor tyrosine kinase, in symptomatic patients with non-small cell lung cancer: a randomized trial. JAMA 2003; 290: 2149–2158.[Abstract/Free Full Text]

9. Miller VA, Kris MG, Shah N et al. Bronchioloalveolar pathologic subtype and smoking history predict sensitivity to gefitinib in advanced non-small-cell lung cancer. J Clin Oncol 2004; 22: 1103–1109.[Abstract/Free Full Text]

10. Mendelsohn J, Baselga J. The EGF receptor family as targets for cancer therapy. Oncogene 2000; 19: 6550–6565.[CrossRef][ISI][Medline]

11. Rusch V, Klimstra D, Venkatraman E et al. Overexpression of the epidermal growth factor receptor and its ligand transforming growth factor alpha is frequent in resectable non-small cell lung cancer but does not predict tumor progression. Clin Cancer Res 1997; 3: 515–522.[Abstract/Free Full Text]

12. Franklin WA, Veve R, Hirsch FR et al. Epidermal growth factor receptor family in lung cancer and premalignancy. Semin Oncol 2002; 29: 3–14.

13. Mendelsohn J, Baselga J. Status of epidermal growth factor receptor antagonists in the biology and treatment of cancer. J Clin Oncol 2003; 21: 2787–2799.[Abstract/Free Full Text]

14. Cappuzzo F, Gregore V, Rossi E et al. Gefitinib in pretreated non-small-cell lung cancer (NSCLC): analysis of efficacy and correlation with HER2 and epidermal growth factor receptor expression in locally advanced or metastatic NSCLC. J Clin Oncol 2003; 21: 2658–2663.[Abstract/Free Full Text]

15. Cappuzzo F, Hirsch FR, Rossi E et al. Epidermal growth factor receptor gene and protein and gefitinib sensitivity in non-small-cell lung cancer. J Natl Cancer Inst 2005; 97: 643–655.[Abstract/Free Full Text]

16. Tsao MS, Sakurada A, Cutz JC et al. Ertotinib in lung cancer – molecular and clinical predictors of outcome. N Engl J Med 2005; 353: 133–144.[Abstract/Free Full Text]

17. Fukuoka M, Yano S, Giaccone G et al. Multi-institutional randomized phase II trial of gefitinib for previously treated patients with advanced non-small-cell lung cancer (The IDEAL 1 Trial) [corrected]. J Clin Oncol 2003; 21: 2237–2246.[Abstract/Free Full Text]

18. Brognard J, Clark AS, Ni Y et al. Akt/protein kinase B is constitutively active in non-small cell lung cancer cells and promotes cellular survival and resistance to chemotherapy and radiation. Cancer Res 2001; 61: 3986–3997.[Abstract/Free Full Text]

19. Mukohara T, Kudoh S, Yamauchi S et al. Expression of epidermal growth factor receptor (EGFR) and downstream-activated peptides in surgically excised non-small-cell lung cancer (NSCLC). Lung Cancer 2003; 41: 123–130.[ISI][Medline]

20. Sansal I, Sellers WR. The biology and clinical relevance of the PTEN tumor suppressor pathway. J Clin Oncol 2004; 22: 2954–2963.[Abstract/Free Full Text]

21. Soria JC, Lee HY, Lee JI et al. Lack of PTEN expression in non-small cell lung cancer could be related to promoter methylation. Clin Cancer Res 2002; 8: 1178–1184.[Abstract/Free Full Text]

22. D'Amico TA, Massey M, Herndon JE et al. A biologic risk model for stage I lung cancer: immunohistochemical analysis of 408 patients with the use of ten molecular markers. J Thorac Cardiovasc Surg 1999; 117: 736–743.[Abstract/Free Full Text]

23. Shiba M, Kohno H, Kakizawa K et al. Ki-67 immunostaining and other prognostic factors including tobacco smoking in patients with resected nonsmall cell lung carcinoma. Cancer 2000; 89: 1457–1465.[CrossRef][ISI][Medline]

24. Lloyd RV, Erickson LA, Jin L et al. p27kip1: a multifunctional cyclin-dependent kinase inhibitor with prognostic significance in human cancers. Am J Pathol 1999; 154: 313–323.[Abstract/Free Full Text]

25. Hayashi H, Ogawa N, Ishiwa N et al. High cyclin E and low p27/Kip1 expressions are potentially poor prognostic factors in lung adenocarcinoma patients. Lung Cancer 2001; 34: 59–65.[ISI][Medline]

26. Hommura F, Dosaka-Akita H, Mishina T et al. Prognostic significance of p27KIP1 protein and ki-67 growth fraction in non-small cell lung cancers. Clin Cancer Res 2000; 6: 4073–4081.[Abstract/Free Full Text]

27. Kolquist KA, Ellisen LW, Counter CM et al. Expression of TERT in early premalignant lesions and a subset of cells in normal tissues. Nat Genet 1998; 19: 182–186.[CrossRef][ISI][Medline]

28. Yang Y, Chen Y, Zhang C et al. Nucleolar localization of hTERT protein is associated with telomerase function. Exp Cell Res 2002; 277: 201–209.[CrossRef][ISI][Medline]

29. Soria JC, Moon C, Wang L et al. Effects of N-(4-hydroxyphenyl)retinamide on hTERT expression in the bronchial epithelium of cigarette smokers. J Natl Cancer Inst 2001; 93: 1257–1263.[Abstract/Free Full Text]

30. Xinarianos G, Scott FM, Liloglou T et al. Telomerase activity in non-small cell lung carcinomas correlates with smoking status. Int J Oncol 1999; 15: 961–965.[ISI][Medline]

31. Maida Y, Kyo S, Kanaya T et al. Direct activation of telomerase by EGF through Ets-mediated transactivation of TERT via MAP kinase signaling pathway. Oncogene 2002; 21: 4071–4079.[CrossRef][ISI][Medline]

32. Smith LL, Coller HA, Roberts JM. Telomerase modulates expression of growth-controlling genes and enhances cell proliferation. Nat Cell Biol 2003; 5: 474–479.[CrossRef][ISI][Medline]

33. Cappuzzo F, Magrini E, Ceresoli GL et al. Akt phosphorylation and gefitinib efficacy in patients with advanced non-small-cell lung cancer. J Natl Cancer Inst 2004; 96: 1133–1141.[Abstract/Free Full Text]

34. Goldstein NS, Armin M. Epidermal growth factor receptor immunohistochemical reactivity in patients with American Joint Committee on Cancer stage IV colon adenocarcinoma: implications for a standardized scoring system. Cancer 2001; 92: 1331–1346.[CrossRef][ISI][Medline]

35. Lantuejoul S, Soria JC, Moro-Sibilot D et al. Differential expression of telomerase reverse transcriptase (hTERT) in lung tumours. Br J Cancer 2004; 90: 1222–1229.[CrossRef][ISI][Medline]

36. SAS Institute Inc. SAS/STAT Software. Changes and Enhancements. Release 8.2. Cary, NC: SAS Publishing 2001.

37. Sturn A, Quackenbush J, Trajanoski Z. Genesis: cluster analysis of microarray data. Bioinformatics 2002; 18: 207–208.[Abstract/Free Full Text]

38. Shigematsu H, Lin L, Takahashi T et al. Clinical and biological features associated with epidermal growth factor receptor gene mutations in lung cancers. J Natl Cancer Inst 2005; 97: 339–346.[Abstract/Free Full Text]

39. Lynch TJ, Bell DW, Sordella R et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med 2004; 350: 2129–2139.[Abstract/Free Full Text]

40. Paez JG, Janne PA, Lee JC et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004; 304: 1497–1500.[Abstract/Free Full Text]

41. Pao W, Miller V, Zakowski M et al. EGF receptor gene mutations are common in lung cancers from "never smokers" and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci USA 2004; 101: 13306–13311.[Abstract/Free Full Text]

42. West KA, Brognard J, Clark AS et al. Rapid Akt activation by nicotine and a tobacco carcinogen modulates the phenotype of normal human airway epithelial cells. J Clin Invest 2003; 111: 81–90.[Abstract/Free Full Text]

43. Tsao AS, McDonnell T, Lam S et al. Increased phospho-AKT (Ser(473)) expression in bronchial dysplasia: implications for lung cancer prevention studies. Cancer Epidemiol Biomarkers Prev 2003; 12: 660–664.[Abstract/Free Full Text]

44. Di Cristofano A, Pandolfi PP. The multiple roles of PTEN in tumor suppression. Cell 2000; 100: 387–390.[CrossRef][ISI][Medline]

45. Torres J, Navarro S, Rogla I et al. Heterogeneous lack of expression of the tumour suppressor PTEN protein in human neoplastic tissues. Eur J Cancer 2001; 37: 114–121.[CrossRef][ISI][Medline]

46. Haga Y, Hiroshima K, Iyoda A et al. Ki-67 expression and prognosis for smokers with resected stage I non-small cell lung cancer. Ann Thorac Surg 2003; 75: 1727–1732.[Abstract/Free Full Text]

47. Albanell J, Lonardo F, Rusch V et al. High telomerase activity in primary lung cancers: association with increased cell proliferation rates and advanced pathologic stage. J Natl Cancer Inst 1997; 89: 1609–1615.[Abstract/Free Full Text]

48. Wang L, Soria JC, Chang YS et al. Association of a functional tandem repeats in the downstream of human telomerase gene and lung cancer. Oncogene 2003; 22: 7123–7129.[CrossRef][ISI][Medline]

49. Mustonen R, Schoket B, Hemminki K. Smoking-related DNA adducts: 32P-postlabeling analysis of 7-methylguanine in human bronchial and lymphocyte DNA. Carcinogenesis 1993; 14: 151–154.[Abstract]

50. Yashima K, Litzky LA, Kaiser L et al. Telomerase expression in respiratory epithelium during the multistage pathogenesis of lung carcinomas. Cancer Res 1997; 57: 2373–2377.[Abstract]





This Article
Abstract
Full Text (PDF)
All Versions of this Article:
16/12/1906    most recent
mdi408v1
E-letters: Submit a response
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Disclaimer
Request Permissions
Google Scholar
Articles by Dutu, T.
Articles by Soria, J.-C.
PubMed
PubMed Citation
Articles by Dutu, T.
Articles by Soria, J.-C.