Lung Cancer 70 (2010) 43 50. Contents lists available at ScienceDirect. Lung Cancer. journal homepage: www.elsevier.

Lung Cancer 70 (2010) 43 50 Contents lists available at ScienceDirect Lung Cancer journal homepage: www.elsevier.com/locate/lungcan Tissue and serum EGFR as prognostic factors in malignant pleural mesothelioma Rabab Gaafar a, Abeer Bahnassy b,, Ibrahim Abdelsalam c, Mahmoud M. Kamel d, Amani Helal a, Amany Abdel-Hamid b, Nelly Aly Eldin e, Nadia Mokhtar b a Medical Oncology Department, National Cancer Institute, Cairo University, Egypt b Pathology Department, National Cancer Institute, Cairo University, Kaser El-Aini Str., Fom El-Khaleg, Cairo 11967, Egypt c Cancer Biology Department, National Cancer Institute, Cairo University, Egypt d Clinical Pathology Department, National Cancer Institute, Cairo University, Egypt e Epidemiology and Biostatistics Department, National Cancer Institute, Cairo University, Egypt article info abstract Article history: Received 18 August 2009 Received in revised form 15 December 2009 Accepted 4 January 2010 Keywords: Malignant pleural mesothelioma Serum and tissue EGFR Prognosis Background: Malignant pleural mesothelioma (MPM) is an asbestos related aggressive tumor. Asbestos causes genetic modifications and cell signaling events that favor resistance to chemotherapy. A variety of receptor tyrosine kinases have been identified to play a central role in various aspects of tumorigenesis. Epidermal growth factor receptor (EGFR) is overexpressed in a variety of epithelial malignancies including lung cancer in which EGFR aberrations not only predict response to EGFR tyrosine kinase inhibitors but also indicate tumor progression. However in MPM, the role of EGFR is less clear. This study was designed to identify serum and tissue EGFR levels in patients with MPM and to evaluate the relationship between serum and tissue EGFR levels and clinicao-pathological prognostic factors and survival. Methods: We investigated 71 cases of MPM for EGFR expression in tissue. Serum EGFR was assessed in 40 out of those 71 cases and 20 healthy subjects as a control. Pre-treatment serum EGFR levels were measured using quantitative enzyme-linked immunosorbent assay. Tissue EGFR protein overexpression was assessed by immunohistochemistry and gene amplification was assessed by the chromogen in situ hybridization (CISH) technique. Results were correlated with the clinical pathological factors of the patients and overall survival (OS). Results: Out of the 71 patients included in the study, 19 had undergone extrapleural pneumonectomy. As for the rest of the patients, 46 received chemotherapy while 6 had only best supportive care. EGFR immuno-reactivity was detected in 74.6% of the cases, 37 (52.1%) cases were positive for EGFR gene amplification by CISH, 31 of them revealed moderate to high (++, +++) EGFR immuno-reactivity. Elevated serum EGFR >2.5 ng/ml (the median concentration of EGFR in MPM) was reported in 45% of the cases. The overall response rate (RR) for the 46 treated patients who received chemotherapy was 24.1%. After a median follow up of 29 months, the median overall survival (OS) was 10 months. Elevated serum and tissue EGFR is significantly associated with advanced disease stage. However neither EGFR overexpression in tissues nor high serum levels were associated with survival rates. Conclusions: EGFR expression is a common feature in MPM patients. High pre-treatment levels of serum EGFR are associated with advanced stage but not with reduced OS. Detailed mutational analysis of EGFR on a larger number of patients is still needed to clarify the exact role of EGFR in MPM patients. 2010 Elsevier Ireland Ltd. All rights reserved. 1. Introduction Malignant pleural mesothelioma (MPM) is an asbestos related tumor with an increasing incidence, which is difficult to detect early and treat effectively. It has been demonstrated that in the next 35 years about one-quarter of a million deaths will occur as a result of this disease in Western Europe [1,2]. In Egypt, 635 cases of MPM were diagnosed at the National Cancer Institute (NCI), Cairo Uni- Corresponding author. Tel.: +20 1 01720698/2 23664649; fax: +20 2 23644720. E-mail address: chaya2000@hotmail.com (A. Bahnassy). versity and in Abassia Chest Hospital in the first four years of the third millennium (2000 2003). Moreover, NCI hospital-based registry showed an increase in the relative frequency of MPM from 0.47 during 2001 to 1.3 during 2003 [3]. Despite recent improvements in the therapeutic management of MPM through surgery, chemotherapy, radiotherapy, and supportive measures, it still has a poor prognosis with a median overall survival ranging from 6 to 12 months only [4]. Several clinicopathologic prognostic scoring systems have been proposed, but they are not able to predict accurately the individual patient s outcome, and therefore specific therapy is possibly deferred. Accordingly, there is an increasing demand for identifying new biological prognostic 0169-5002/$ see front matter 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.lungcan.2010.01.002

44 R. Gaafar et al. / Lung Cancer 70 (2010) 43 50 markers as well as for the development of novel effective therapies [5]. The epidermal growth factor receptor (EGFR) is a member of the family of EGF-related tyrosine kinase (TK) receptors. Upon ligands binding, the receptors homo- or hetero-dimerize. Subsequently, it activates receptors intrinsic TK activity and broad downstream signaling cascades, mainly including Ras-Raf-MAP-kinase pathway, PI3K-Akt pathway and STAT pathway leading to strong stimulatory effect on cell proliferation, differentiation, survival, angiogenesis, and migration [6 8]. The EGFR is expressed in a variety of human epithelial tumors including lung, head and neck, colorectal, and breast cancers [9]. EGFR has emerged as a critical tumorigenic factor in the development and progression of non small cell lung cancer (NSCLC). Two specific EGFR tyrosine kinase inhibitors (TKIs), gefitinib (ZD, 1839, Iressa) and erlotinib (OSI-774, Tarceva), have been developed and used clinically in the treatment of advanced NSCLC [10 12]. Both agents can induce dramatic clinical response in patients who fail chemotherapy by disrupting EGFR signaling via competing with ATP for the binding sites at the TK domain, thus inhibiting the phosphorylation and activation of the EGFRs and their downstream signaling network [10]. On the other hand data regarding EGFR in MPM are still immature and show controversial results. So, EGFR expression levels ranging from 0% to 97% have been reported in MPM patients with similar variability in patients outcome [13 15]. In most of these studies EGFR aberrations were evaluated in tissues either by immunohistochemical staining or by studying genetic mutations. In a recent study, Destro et al. [15] demonstrated a significant correlation between EGFR immunohistochemical expression and the corresponding mrna levels by real time PCR. In this work, EGFR was not expressed in any of normal pleural samples. This data confirms and enlarges previous observations suggesting an important role of EGFR overexpression in pleural carcinogenesis. EGFR is also readily identifiable and quantifiable in serum and several reports have indicated that changes in the serum levels of EGFR are associated with aggressive cancer development, in kidney, gastric and lung cancer [11,16,17]. However, the usefulness of identifying EGFR level in serum from MPM patients is not yet defined. In this study we investigated the potential prognostic and predictive values of elevated tissue and serum EGFR expression as well as EGFR gene amplification in MPM patients through correlation with the standard clinico-pathological features of patients and overall survival rates. 2. Patients and methods 2.1. Tumor samples In this prospective study, 71 MPM patients who attended the National Cancer Institute (NCI), Cairo University, Egypt during the period from 2004 to 2007 were investigated for EGFR expression in tissues. Pre-treatment serum samples were available for 40 out of those 71 patients. Paraffin blocks for all patients were recruited from the archive of the pathology department following diagnosis whereas pre-treatment serum samples were collected from patients at presentation and kept at 70 C. In addition, serum samples were obtained from 20 healthy volunteers and used as a control for serum EGFR. All cases were diagnosed as MPM and classified using the WHO criteria on hematoxylin and eosin-stained sections combined with immunohistochemistry, using the routine diagnostic panel for MPM (Table 1) [18]. A written consent should be obtained from all patients prior to enrolment in the study, and the ethical committee of the NCI approved the protocol which was Table 1 Clinical features of the 71 malignant pleural mesothelioma cases studied. Variable Number of patients (%) Age 46 years 36 (51) >46 years 35 (49) Median age 46 Gender Male 45 (63.4) Female 26 (36.6) Male:female ratio 1.7:1 Asbestos exposure Positive 50 (70.4) Negative 21 (29.6) Performance status 1 41 (57.8) 2 23 (32.4) 3 7 (9.8) Histopatholgical type Epithelial 39 (54.9) Mixed 19 (26.8) Sarcomatoid 7 (9.8) NOS 6 (8.5) Tumor stage (IMIG) a Early (I and II) 19 (26.8) Advanced (III and IV) 52 (73.2) T-stage Early I 7 (9.85) II 27 (38.03) Advanced III 20 (28.17) IV 17 (23.9) Clinical manifestation Dyspnea 54 (76.1) Chest pain 48 (67.6) Pleural effusion 47 (66.2) Cough 33 (46.5) a MIG: International Mesothelioma Interested Group. in accordance with the ethical guidelines of the 1975 Declaration of Helsinki. 2.2. Clinical workup and treatment protocols According to IMIG staging classification [19], patients were divided into early (stages I and II) and advanced (stages III and IV) stages and accordingly, they received treatment in the form of extrapleural pneumonectomy or chemotherapy, respectively. The 19 operable cases were pathologically staged whereas the remaining 52 cases were clinically staged. Advanced cases received platinum based chemotherapy with or without radiotherapy according to the treatment protocols of the NCI, Cairo. 2.3. Evaluation of serum EGFR expression After collection of the venous blood samples from patients (40 samples) and control (20 samples) subjects, the samples were centrifuged at 3000 rpm for 10 min and then stored at 70 C until assay. The human Active EGFR ELISA (Bender Medysystems Diagnostics GmbH, Rennweg 95b A-1030 Vienna, Austria) kit was used for the detection of quantitative level of EGFR in serum according to the recommendation of the manufacturer.

R. Gaafar et al. / Lung Cancer 70 (2010) 43 50 45 2.4. Immunohistochemistry Immunohistochemistry was performed to confirm the diagnosis of MPM and to assess the expression of EGFR protein using the following panel: calretinin, keratin 5/6, epithelial membrane antigen and/or cytokeratin, CEA, vimentin, and the EGFR pharmdxtm staining kit (all from Dako, USA) according to manufacturer s instructions. The standard streptavidin biotin peroxidase detection technique was used as previously described [18]. The antigen retrieval method was performed by microwave pre-treatment in 0.01 M citrate buffer. The primary antibody was applied and incubated overnight at 4 C in a humidified chamber. After 3 washes in PBS, the secondary antibody and the avidin biotin complex (ABC) were applied to slides. Diaminobenzidine (DAB) was used as a chromogen and sections were counterstained using Mayer s hematoxylin. Manufacturer s protocol for EGFR staining was strictly followed. 2.7. Statistical analysis Univariate and multivariate survival analyses were done to evaluate the prognostic significance of clinico-pathological factors and EGFR expression on the overall survival. Specifically, the product limit or Kaplan-Meier method was applied as univariate analysis to estimate the survival probabilities, and the log-rank test (LT) was performed to determine the statistical differences between categories of each prognostic factor. Cox s proportional hazards regression model was performed as multivariate analysis to estimate the role of each prognostic factor on survival outcome, adjusted for the possible confounding effect of statistically significant covariates included in the same model. Asymptotic standard errors of coefficients were used to calculate 95% confidence intervals (95% CI) of hazard ratio (HR; relative risk). Chi-square test ( 2 ) was used to test the association variables for categorical data. Survival data were analyzed using the EGRET statistical software package. 2.5. Interpretation of immunohistochemical results Assessment of EGFR was based on a membranous staining pattern. Absence of specific membranous staining within tumor cells was considered negative. In case of positivity (defined as any immuno-staining of tumor cell membranes), the intensity of staining, and the percent of positive cells were assessed according to the instructions of the EGFR pharmdxtm staining kit and reported (1%, 5%, 10%, etc.). 2.6. Chromogen in situ hybridization (CISH) Paraffin-embedded tissue sections were de-paraffinized in two changes of xylene for 5 min each; xylene was removed in three washes of ethanol for 3 min each (100%, 100%, and 95%) and the slides were washed in distilled running water for 5 min. Slides were placed in heated (>90 C) CISH Pre-treatment Buffer (Zymed Laboratories) and microwaved on high power for 30 min, then rinsed in distilled water for 5 min at room temperature (RT). The tissue was digested for 10 min with pepsin digestion solution (Zymed Laboratories) at RT, washed twice in distilled water for 5 min each, dehydrated in 90%, 95%, and 100% alcohol for 2 min each, and then dried in a 37 C oven. Five seven micro-liters of EGFR Amplification Probe (Zymed-Invitrogen) was applied to the designated area and a cover slip was applied and sealed with rubber cement. Slide was dried at 37 C, followed by probe denaturation at 95 C for 5 min and hybridization at 37 C overnight. The slides were then washed in 0.5% SSC for 5 min at RT, followed by 0.5% SSC for 5 min at 75 C, and water for 5 min at RT. For immuno detection, slides were placed in 3% hydrogen peroxide in absolute methanol for 10 min, washed in phosphate-buffered saline with 0.025% Tween 20 (PBST) three times for 2 min each, incubated with nonspecific blocking solution (Zymed-Invitrogen) for 10 min at RT and then with mouse anti-digoxigenin antibody for 30 min at RT. Slides were then washed in PBST twice for 2 min each, incubated with horseradish peroxidase-conjugated goat antimouse antibody for 15 min, and finally washed again with PBST twice for 2 min each. The slides were then incubated with DAB chromagen (Zymed-Invitrogen) for 30 min at room temperature and washed in distilled water twice for 2 min each. Slides were counterstained with hematoxylin [20]. The CISH-prepared slides were examined at X400 by bright field microscopy. As in prior studies [20], 200 tumor nuclei were examined in each case. In the cases where the number of signals per tumor nucleus varied between tumor cells, the range of signals counted was recorded for each case. 3. Results The characteristics of the 71 MPM patients enrolled in this are shown in Table 1. The median age for the patients was 46 years (range 26 77 years). Forty-five patients were males and 26 were females with a male to female ratio of 1.7:1. A history of asbestos exposure was given in 50 (70.4%) patients. Dyspnea was the commonest presentation being encountered in 54 patients, followed by chest pain in 48 patients, pleural effusion in 47 patients and cough in 33 patients. The Eastern Cooperative Oncology Group (ECOG) PS was 1 in 41 (57.7%) patients and >1 in 30 (42.3%) patients. Fiftytwo (73.2%) patients were diagnosed as having advanced disease stage (III and IV) at presentation whereas 19 patients (26.8%) presented at an early stage (I and II). Extrapleural pneumonectomy was performed for the 19 early stage cases. Out of the 52 patients with advanced disease stage, 46 were treated with platinum based chemotherapy and/or radiotherapy according to NCI, Cairo protocols whereas 6 were treated with palliative measures only. Two patients out of 46 (4%) achieved complete remission (CR) and 9 (20%) had partial remission (PR); the overall response rate was 24%. On the other hand 22 (48%) patients had stable disease and 13 (28%) show disease progression. After a median follow up of 29 months, the median survival for the early cases 14.3 months, while the median survival for the advanced cases was 10 months. 3.1. Evaluation of serum EGFR Positive or negative serum EGFR was determined using a cutoff value of 0.23 ng/ml, which represents the median concentration of the control group. The median concentration for the 40 cases assessed was 2.5 ng/ml. It was significantly higher than the median concentration for normal control subjects (0.23 ng/ml; p = 0.001). Eighteen patients (45%) had elevated serum EGFR more than 2.5 ng/ml (Fig. 1). A statistically significant correlation was found between a high serum EGFR expression (>2.5 ng/ml) and advanced disease stage (p = 0.05) as well as between serum EGFR level and T-stage (p = 0.001). No relation was found with other clinico-pathological features of the patients, asbestos exposure, and response to treatment or OS (Table 2). However, using 0.23 ng/ml as a cutoff there was a trend for better response to treatment in patients with low serum EGFR levels (p = 0.09) as 53% (9/17) of the patients with good response to treatment have a high serum EGFR >0.23 ng/ml compared to 87.5% (7/8) of the patients with aggressive disease and a high serum EGFR >0.23 ng/ml.

46 R. Gaafar et al. / Lung Cancer 70 (2010) 43 50 Fig. 2. A case of epithelioid malignant pleural mesothelioma showing positive membrano/cytoplasmic immuno-staining (+++) for EGFR. 3.2. Evaluation of tissue EGFR Fig. 1. Epidermal growth factor receptor (EGFR) serum levels in malignant pleural mesothelioma cases. The median concentration for the control group was 0 23 ng/ml (range 0 0.96) and the median concentration for the tested group was 2.5 ng/ml (range 0.5 7.6, p = 0.001). Table 2 Serum EGFR concentration in relation to the clinicopathologic features of the studied patients. Variables Serum concentration 2.5 (22) Serum concentration > 2.5 (18) p value Mean age (years) 46 13 (59.1) 10 (55.6) >46 9 (40.9) 8 (44.4) 0.92 Sex Male 14 (63.6) 10 (55.6) Female 8 (36.4) 8 (44.4) 0.6 EGFR immuno-reactivity was detected in 53/71 cases (74.6%, Fig. 2), 22 (30.98%) were scored (+), 21 (29.6%) were scored (++), 10 (14.1%) were scored (+++). No immuno-reactivity was detected in normal pleural samples. The ranges of CISH signals recorded in the analyzed cases were as follows: in the majority of the nonamplified cases, the number of CISH signals per cell ranged from 2 to 5. In gene-amplified cases, the number of CISH signals per cell was estimated at 5 30 with at least focal clustering of signals consistent with tandem gene repeats (Fig. 3). Out of the 71 cases examined, 37 (52.1%) cases were positive for EGFR gene amplification by CISH, 31 of them revealed moderate to high (++, +++) EGFR immuno-reactivity and 6 revealed mild (+) EGFR immunoreactivity. The concordance between EGFR protein expression and gene amplification (86%) was statistically significant (p = 0.01). EGFR overexpression and gene amplification were significantly associated with T-stage of the tumors (p = 0.021) but not with any of the clinico-pathological factors of studied patients, asbestos exposure or OS (Table 3 and Fig. 3). Using 0.23 ng/ml as a cutoff point, there was a statistically significant relation between increased EGFR protein expression and/or gene amplification and elevated EGFR serum levels (p < 0.05). However, there was no statistically significant relation between EGFR serum levels and neither protein overexpression nor gene amplification (p = 0.0671 and p = 0.884) using 2.5 ng/ml as a cutoff point. Performance status <2 14 (63.6) 10 (55.6) 0.6 2 8 (36.4) 8 (44.4) Asbestos exposure Negative 9 (40.9) 9 (50) Positive 13 (59.1) 9 (50) 0.56 Histopathologic subtype Epithelioid 13 (59.1) 13 (72.2) Sarcomatoid 2 (9.1) 1 (5.6) Mixed 4 (18.2) 3 (16.6) NOS 3 (13.6) 1 (5.6) 0.3 Tumor stage Early (I and II) 10 (45.5) 3 (16.7) Advanced (III and IV) 12 (54.5) 15 (83.3) 0.05 T-stage I and II 18 (81.8) 2 (11.1) III and IV 8 (17.2) 16 (88.9) 0.001 Pleural effusion Negative 13 (59.1) 8 (44.4) Positive 9 (40.9) 10 (55.6) 0.57 Response to CTH Clinical benefit 10 (66.7) 7 (70) 0.8 a Progression 5 (33.3) 3 (30) a Serum samples were available for 25 out of the 46 treated patients. Fig. 3. A case of epithelioid malignant pleural mesothelioma showing multiple brown nuclear dots and clusters denoting EGFR gene amplification as measured by chromogen in situ hybridization (CISH) technique.

R. Gaafar et al. / Lung Cancer 70 (2010) 43 50 47 Table 3 Tissue EGFR expression in relation to the clinico-pathologic features of the 71 patients studied. EGFR negative (18) EGFR positive (53) p value Mean age (years) 46 8 (44.5) 28 (52.8) >46 10 (55.5) 25 (47.2) 0.84 Sex Male 9 (50) 36 (67.9) Female 9 (50) 17 (32.1) 0.92 Performance status <2 11 (61.1) 30 (56.6) 0.54 2 7 (38.9) 23 (43.4) Asbestos exposure Negative 8 (44.5) 13 (24.5) Positive 10 (55.5) 40 (75.5) 0.78 Histopathologic subtype Epithelioid 8 (44.4) 31 (58.5) Sarcomatoid 2 (11.1) 5 (9.5) Mixed 6 (33.4) 13 (24.5) NOS 2 (11.1) 4 (7.5) 0.77 Tumor stage Early (I and II) 4 (22.2) 15 (28.3) Late (III and IV) 14 (77.8) 38 (71.7) 0.88 T-stage I and II 14 (77.8) 20 (37.7) III and IV 4 (22.2) 33 (62.3) 0.024 Pleural effusion Negative 6 (33.4) 18 (34) 0.62 Positive 12 (66.6) 35 (66) Response to therapy Clinical benefit 10 (76.9) 23 (70) Progression 3 (23.1) 10 (30) 0.76 3.3. Survival analysis No significant relation was found between EGFR expression in serum or tissues and the OS rate. Similarly, no relation was found between EGFR gene amplification by CISH and OS (Table 4). The median survival for cases having serum EGFR more than 2.5 ng/ml and equal or less than 2.5 ng/ml was 6 months (95% confidence interval [CI], 2 10 months) and 8 months (95% CI, 5 11 months); respectively (p = 0.44, Fig. 4). Patients with EGFR immunopositive tumors had a median survival of 12 months (95% CI, 9 15 months) compared to 8 months (95% CI, 3 13 months) for patients with negative tumors (p = 0.27, Fig. 4). Similarly, patients with EGFR gene amplification had a median survival of 10 months (95% CI, 8 16 months) compared to 7 months (95% CI, 5 13 months) for patients with negative tumors (p = 0.26, Fig. 4) (Fig. 5). 4. Discussion In lung cancer, EGFR gene mutation at the kinase domain and EGFR gene amplification were reported to be predictors of the response to EGFR tyrosine kinase inhibitors [13]. Therefore, some anti temporal drugs have been developed against EGFR. In MPM, data regarding EGFR expression and its prognostic significance are still scarce and immature. In the current study, EGFR serum levels were significantly higher in MPM patients than healthy controls (2.5 vs 0.23 ng/ml, p = 0.001). However, this elevated serum level has no prognostic impact since no difference in the OS rates was found between patients with low serum EGFR (<2.5 ng/ml) and those with high serum EGFR levels ( 2.5 ng/ml). However, a high serum EGFR was significantly associated with advanced disease stage since 23.1% (3/13) of patients with early stage had serum EGFR >2.5 ng/ml compared to 55.6% (15/27) in late stage (p = 0.05). Moreover, there was a significant correlation between increased EGFR levels, both in serum and tissues, as well as EGFR gene amplification and T-stage, which is considered a surrogate prognostic factor in MPM. Our results are comparable to those of Betta et al. [21] and Schneider et al. [22] who found no differences in serum values of EGFR between MPM patients with varied clinical outcomes and attributed this to the small number of cases studied. In the present study, there was a trend for patients with low serum EGFR levels to show better response to treatment compared to those with high serum EGFR >0.23 ng/dl as a cutoff value separating patients from healthy controls though the difference did not reach a statistically significant value (p = 0.09). We also assessed whether tissue EGFR protein expression and/or gene amplification have any prognostic impact in our studied group with MPM. Seventy-five percent of our cases revealed EGFR immuno-reactivity and 69% revealed EGFR gene amplification by CISH technique (89% concordance). However, no significant difference was reported between high EGFR protein expression or EGFR gene amplification and neither patient response nor OS. Our reported frequency of tissue EGFR protein expression (in tissues) and gene amplification are higher than those of Okuda et al. [13] who studied EGFR gene mutation, amplification and protein expression in 25 MPM Japanese patients and found that only 8/25 (32%) cases were positive for the EGFR protein; all of them were epithelioid. Fluorescence in situ hybridization (FISH) analysis revealed three low polysomy and one high polysomy case only; all of them expressed EGFR protein. The low polysomy cases included one biphasic type and two epithelial types, and the high polysomy case was epithelial type. Our study dose not show any significant relation between EGFR gene amplification or EGFR protein expression and the histopathology type of tumor. Table 4 Cox regression analysis results of EGFR in serum and tissue in relation to the overall survival of the studied cases. Median survival B SE p value OR 95.0% CI for OR Lower Upper Serum EGFR <2.5 ng/ml 8 3.557 0.996 0.44 15.864 5 11 >2.5 ng/ml 6 2 10 Tissue EGFR (protein) Negative 8 5.998 1.89 0.27 24.511 3 13 Overexpression 12 9 15 EGFR gene amplification Negative 7 2.116 3.974 0.026 8.749 5 13 Positive 10 8 16

48 R. Gaafar et al. / Lung Cancer 70 (2010) 43 50 Fig. 4. EGFR expression in serum and tissues in relation to the history of asbestos exposure (p = 0.56 and p = 0.78 respectively) and disease stage according to IMIG (p = 0.05 and p = 0.88 respectively) in malignant pleural mesothelioma patients studied. Our frequency of EGFR protein expression is also higher than, Edwards et al. [2], Govindan et al. [23] who reported EGFR protein expression in 44% and 58% of their studied cases (21) respectively. On the contrary, Garland et al. [24] reported EGFR protein expression in 88% of MPM cases and Govindan et al. [25] reported EGFR overexpression in 97% of patients undergoing therapy with the EGFR TKI gefitinib, as part of a phase II study. This heterogeneity in the results of different studies could be explained by differences in sample size, patients characteristics, the detection method, the antibody used to determine tissue EGFR protein expression, the difference in interpreting the results of IHC including the intensity of expression and the localization of the receptor where some authors reported only membranous staining as opposed to cytoplasmic staining, while others did not report any preferential localization of the receptor. No significant relation was found in the current study between EGFR protein expression or gene amplification and any of the standard clinico-pathologic prognostic factors (except the stage of the tumor) or response to therapy. Prior studies have reached variable conclusions regarding the prognostic significance of EGFR in MPM. Some authors reported that the tumor cell expression of the growth factors was associated with favorable outcome in MPM [2,26]. Others [27] demonstrated that EGFR is a poor prognostic factor. The inverse correlation between EGFR and survival may be explained partly by the differences in expression between the epithelioid and sarcomatoid cell types, since Dazzi et al. [28] found that, after the cell type was taken into consideration, lack of EGFR immuno-staining was no longer significant. Our findings agree with those of Govindan et al., Okuda et al., and Destro et al. [23,13,15] that EGFR expression and/or gene amplification in tissue are not related to patients survival or response to the available therapeutic modalities. Although EGFR status did not correlate with overall survival, its frequency in MPM suggests that it

R. Gaafar et al. / Lung Cancer 70 (2010) 43 50 49 Fig. 5. Kaplan and Mayer curves showing the overall survival rates in patients with elevated serum and tissue EGFR expression in malignant pleural mesothelioma patients. may be a novel therapeutic target. Confirmation of this hypothesis comes from in vitro studies where treatment of MPM cell lines with the selective EGFR TKI (Gefitinib) led to significant dose dependent reduction of colony formation of cells grown in soft agarose [29]. Moreover, the EGFR TKI PD153035 inhibited motility and invasion of MPM cells in a matrigel assay [30]. However, results from phase II studies showed that EGFR tyrosine kinase inhibitors did not appear to be effective in unselected patients. The median overall survival time was 4 10 months and one-year survival was estimated at 43%, although some patients had objective response or stable disease [14,25,26]. In addition, there is no randomized data from phase III studies examining measurable clinical benefits of EGFR pathway inhibitors in MPM patients. Therefore, further studies are required to evaluate the role of targeted therapy in patients with activated EGFR pathways. We conclude that increased EGFR protein expression (in serum and tissues) as well as EGFR gene amplification are common in Egyptian patients with MPM indicating that EGFR may be a target for selective therapies in a subset of patients. However, EGFR was not found to be of prognostic value in our cases since no association was found between EGFR protein expression in serum and tissues or EGFR gene amplification and OS, patients response to treatment or the standard clinico-pathological prognostic factors except for advanced disease stage. A more extended study including large number of patients is warranted to define detailed genetic changes affecting EGFR and other family members. Conflict of interest Conflict of interest is declared. References [1] Peto J, Decarli A, La Vecchia C, et al. The European mesothelioma epidemic. Br J Cancer 1999;79:666 72. [2] Edwards J, Swinson D, Jones J, Waller D, O Byrne K. EGFR expression: associations with outcome and clinicopathological variables in malignant pleural mesothelioma. Lung Cancer 2006;54:399 407. [3] Gaafar R, Aly Eldin N. Epidemic of mesothelioma in Egypt. Lung Cancer 2005;49:17 20. [4] Edwards J, Abrams K, Leverment J, Spyt T, Waller D, O Byrne K. Prognostic factors for malignant mesothelioma in 142 patients: validation of CALGB and EORTC prognostic scoring systems. Thorax 2000;55:731 5. [5] Ogino H, Yano S, Kakiuchi S, Yamada T, Ikuta K, Nakataki E, et al. Novel dual targeting strategy with vandetanib induces tumor cell apoptosis and inhibits angiogenesis in malignant pleural mesothelioma cells expressing RET oncogenic rearrangement. Cancer Lett 2008;265(1):55 66. [6] Olayioye M, Neve R, Lane H, Hynes N. The ErbB signaling network: receptor heterodimerization in development and cancer. EMBO J 2000;19:3159 67.