Epidermal growth factor receptor mutations in malignant pleural and peritoneal mesothelioma
- Yasunori Enomoto1,
- Takahiko Kasai1,
- Maiko Takeda1,
- Masato Takano1,
- Kohei Morita1,
- Eiji Kadota2,
- Norishige Iizuka2,
- Hiroshi Maruyama3,
- Joji Haratake4,
- Yu Kojima5,
- Naoya Ikeda5,
- Naoki Inatsugi6,
- Akitaka Nonomura1
- 1Department of Diagnostic Pathology, Nara Medical University School of Medicine, Nara, Japan
- 2Department of Laboratory Medicine and Pathology, Kishiwada City Hospital, Osaka, Japan
- 3Department of Pathology, Hoshigaoka Koseinenkin Hospital, Osaka, Japan
- 4Department of Diagnostic Pathology, Saiseikai Yahata General Hospital, Yahata, Fukuoka, Japan
- 5Department of Surgery, Nara Prefectural Mimuro Hospital, Nara, Japan
- 6Department of Surgery, Kenseikai Dongo Hospital, Nara, Japan
- Correspondence to Dr Takahiko Kasai, Department of Diagnostic Pathology, Nara Medical University School of Medicine, Shijo-cho 840, Kashihara, Nara 634-8521, Japan;
- Accepted 26 January 2012
- Published Online First 12 March 2012
Background Epidermal growth factor receptor (EGFR) gene mutation at the kinase domain and EGFR gene amplification are reported to be predictors of the response to EGFR tyrosine kinase inhibitors in lung cancer cases. In malignant mesothelioma (MM), the role of EGFR is less clear.
Methods Thirty-eight MM specimens were submitted to EGFR mutation evaluation, and compared with the results of immunohistochemical staining and fluorescence in situ hybridization (FISH) analysis. DNA was extracted from paraffin blocks and PCR was performed to amplify exon regions 18–21 of the EGFR gene. Direct sequencing of the purified PCR products was performed.
Results Five EGFR missense mutations were detected in six of the 38 patients (16%); two of these mutations were novel, two were originally detected in non-small cell lung carcinoma, and one resembled a location previously noted for malignant peritoneal mesothelioma.
Conclusion As far as the authors are aware there has been no report of the EGFR mutation of MM in Japanese cases, but in this study EGFR missense mutations were detected in some cases. EGFR mutation results were not related to immunohistochemical and FISH analysis.
- Cancer research
- epidermal growth factor receptor
- gastric pathology
- gene copy number gain
- malignant mesothelioma
- malignant tumours
- surgical pathology
Malignant mesothelioma (MM) is an aggressive tumour that develops from the pleura, the peritoneum, or other mesothelial surfaces and is commonly associated with asbestos exposure.
The estimated annual incidence is approximately 3000 individuals in the USA and approximately 1000 cases in Japan.1 2 There is no standard care for MM. The current combination therapy with surgery (extrapleural pneumonectomy), chemotherapy and radiation therapy is only palliative; the median survival ranges from 6 to 18 months. A novel anti-folate, pemetrexed, was recently approved for use with cisplatin in Japan; a phase III study showed that the response rate was 41.3%, time to progression was 5.7 months, and overall survival time was 12.2 months.3 Most of the patients relapsed within a year after the start of treatment. New treatments are thus needed for patients with MM. The poor prognosis, as for other solid carcinomas, is probably related to its complex biology.
Epidermal growth factor receptor (EGFR) is a receptor tyrosine kinase important in transducing extracellular signals from the cell surface to the cell interior, mediating crucial processes such as cell proliferation, differentiation, migration and apoptosis. Dysregulated expression of these receptors can lead to aberrations of homeostatic cellular processes, resulting in malignant transformation of cells. Activating EGFR mutations have been reported in cancers such as non-small-cell lung cancer (NSCLC) and head and neck cancers, and are predictive of the response to gefitinib or erlotinib therapy.4–6 Many mutations in the EGFR gene have been reported in NSCLC but only a few have been confirmed, either from in-vitro studies or from tumour responses in NSCLC patients, to be associated with responses to EGFR tyrosine kinase drugs.4 5 These mutations are usually found in exons 18, 19, 20 and 21, and include missense substitutions such as G719A/S and L858R and deletions like E746 to A750 (removal of amino acids glucine–leucine–arginine–glucine–alanine), which are associated with sensitivity to tyrosine kinase inhibitors (TKI). Mutations associated with resistance to EGFR TKI are D761Y and T790M.7 8
It was reported that immunohistochemical overexpression of the EGFR protein occurs in MM, but gefitinib was not found to be effective in the treatment of the mesothelioma-like lung cancer.9–11 In many previous reports, there seemed to be no EGFR mutation in MM patients. We have reported the results of immunohistochemical and fluorescence in situ hybridization (FISH) analysis for another journal, which is in press now.12
The aim of the present study was to evaluate the EGFR mutation in comparison with gene copy number gain and protein expression in Japanese cases of malignant pleural and peritoneal mesotheliomas.
Materials and methods
Thirty-eight cases of MM (22 pleural, 16 peritoneal; median age 63 years, range 41–81 years) were collected from the archives of the Department of Pathology, Nara Medical University, Kishiwada City Hospital, Hoshigaoka Koseinenkin Hospital, Saiseikai Yahata General Hospital, Kenseikai Dongo Hospital and Nara Prefectural Mimuro Hospital. The histological subtype was epithelioid in 25 cases, biphasic in seven cases and sarcomatoid in six cases (table 2). Ten benign mesothelial lesions (four adenomatoid tumours in the uterus, three benign multicystic mesotheliomas and three mesothelial hyperplasias) and five pleuritis cases were also examined. This work was carried out in accordance with the Declaration of Helsinki (2000) of the World Medical Association and was approved by our institutional review board. All tumour specimens were fixed in 10% neutral formalin and embedded in paraffin, after which 4 μm-thick sections were stained in H&E for histological analysis. Paraffin-embedded tissue blocks were available in all cases. The diagnosis of MM was based on the combination of clinical findings, imaging and gross observations at surgery and on routine H&E histology. The diagnosis was confirmed by immunohistochemical staining (CEA, TTF-1, calretinin, WT-1, D2-40, cytokeratin 5/6, and so on).
For EGFR staining, the EGFR pharmDx assay (K1492; Dako-Cytomation, Tokyo, Japan) was used according to the Bond polymer method using the autoimmunostainer Bond MAX (Leica Microsystems, Wetzler, Germany), as previously reported with modification.13 The antigen retrieval step was performed using Bond epitope retrieval solution 1 (citrate-based pH 6.0 solution). Only the membranous pattern of expression was considered specific. Staining intensity was scored on a 0 to 3+ scale, on the basis of the percentage of positive cells and the intensity of staining (0 for no membranous staining, 1+ for weak staining in <5% of cells, 2+ for moderate staining in 5–50% of cells, and 3+ for strong staining in >50% of cells). Cases scoring 3+ were considered as positive for immunohistochemical analysis.9
Fluorescence in situ hybridization
EGFR gene status evaluation was performed by FISH on 3 μm tissue sections, as previously reported with modification.13 After paraffin sections were depariffinised, dehydrated in ethanol and air-dried, dual-colour FISH analysis was performed using LSI EGFR/CEP7 probes (Abbott, Tokyo, Japan). The LSI EGFR probe is labelled in Spectrum Orange and covers an approximately 300 kb region containing the entire EGFR gene at 7p12. The CEP 7 probe, labelled in Spectrum Green, hybridises to the α satellite DNA located at the centromere of chromosome 7. Pretreatment steps were performed using the histology FISH accessory kit (Dako Cytomation) as follows: The section in pretreatment solution (MES; 2-ethanesulphonic acid buffer) incubated at 105°C for 5 min was digested with pepsin at 37°C for 15 min. The slides were denatured at 75°C for 5 min in 70% formamide (Chemicon, Billerica, Massachusetts, USA) and dehydrated in ethanol. The probes were denatured for 5 min at 75°C before hybridization. Then slides were hybridized overnight at 37°C and washed in 2×SSC/0.3% Igepal at 72°C for 2 min (Sigma, St Louis, Missouri, USA), and were counterstained with DAPI/antifade (Abbott, Tokyo, Japan). The slides were scored on a cell-by-cell basis using a fluorescence microscope (Keyence; HS BZ-9000) and HS BZ-II analysis application equipped with filter sets with single and dual band excitors for Spectrum Green, Spectrum Orange and DAPI (ultraviolet 360 nm). The histological areas previously selected on the H&E-stained sections were identified on the FISH-treated slides. Only individual and well-delineated cells were scored. Overlapping cells were excluded from the analysis. At least 60 cells were scored for each case. Patients were classified according to the criteria proposed by Hirsch et al14 and revised by Martin et al: 15(1) disomy (balanced disomy) for both the EGFR gene and chromosome 7—two copies of chromosome 7 identified in more than 50% of neoplastic cells; (2) low polysomy—three or four copies of chromosome 7 identiﬁed in more than 50% of neoplastic cells; (3) high polysomy—more than four copies of chromosome 7 in more than 50% of neoplastic cells; (4) gene ampliﬁcation—a gene-to-chromosome ratio of more than 3, or finding clusters of ampliﬁcation. The last two were defined as copy number gain (FISH positive).14 15
Genomic DNA was extracted from formalin-fixed, paraffin-embedded tumour sections from each representative case using the QIAamp DNA extraction kit for FFPE formalin fixed paraffin embeded tissues (Qiagen, Hilden, Germany) according to the manufacturer's protocol. Briefly, paraffin was removed using xylene and residual xylene removed with ethanol. Buffer ATL was added to the deparaffinised tissue, heated at 98°C for 15 min and then cooled to room temperature. Proteinase K was added to the tissue and incubated at 56°C for 16 h. Afterwards, the tissue mixture was incubated at 90°C for 1 h and cooled to room temperature. Buffer AL was then added and washing steps were performed according to the manufacturer's instructions. DNA yield and purity were quantitated and assessed using Nanodrop (Thermo Fisher Scientific, Waltham, Massachusetts, USA). PCR was then performed for all DNA samples using primers designed to amplify exons 18, 19, 20 and 21 of the EGFR gene. The primer sequences are shown in table 1. Essentially, each PCR reaction consisted of 1×PCR buffer, 0. 2 mM dNTP, 0. 3 μM forward and reverse primers, 1.25 U Taq DNA polymerase and 250 ng of genomic DNA in a total volume of 50 μl. The PCR cycling programme was as follows: (1) 94°C for 4 min (one cycle); (2) 94°C for 1 min, 60°C for 1 min, 72°C for 1 min (40 cycles); and (3) 72°C for 10 min (one cycle).16 A non-template (DNA) control represented the negative control and was included in every PCR run. PCR products were analysed by performing electrophoresis on a 2% agarose gel stained with ethidium bromide. PCR products were purified using a Qiagen PCR purification kit (Qiagen). The purified PCR amplicons were sequenced by premix sequence analysis (Takara Bio Inc, Shiga, Japan). Sequencing was carried out in both forward and reverse directions. Data analyses of EGFR mutations were performed by Takara Bio Inc. EGFR mutations detected in the initial round of sequencing were confirmed by subsequent rounds of independent PCR and sequencing reactions. Only mutations confirmed by subsequent rounds are reported. In the EGFR mutation-positive cases, we analysed other DNA from another paraffin block in the same case, committed to Mitsubishi Medicine, and so confirmed that the results were not caused by PCR artifacts. Moreover, we also examined the absence of the mutation in background non-neoplastic regions, and so confirmed that they were not somatic mutations.
Fisher's exact probability test was used to compare the positive rates between groups. All p values were adjusted for multiple comparisons using Bonferroni correction at p<0.05.
The EGFR immunohistochemical staining results are shown in tables 2 and 3. No statistical associations were found between EGFR expression, gender, age, or histological subtype. Overall EGFR immunoreactivity was detected in 37 of 38 (97%) cases, while 3+ staining (EGFR immunohistochemical positive) was found in 20 of 38 (53%; figure 1).
By site of origin, immunohistochemical EGFR positive reaction was seen in 11 of 22 (50%) pleural MM, and in 9 of 16 (56%) peritoneal MM.
By histology subtype, immunohistochemical EGFR-positive reaction was demonstrated for 17/24 (67%) epithelioid, 2/7 (28%) biphasic and 1/7 (14%) sarcomatoid MM. An immunohistochemical EGFR-positive reaction was more common in cases with epithelioid than in cases with non-epithelioid histology (67% vs 21%), and the difference was statistically significant (p<0.01).
Thirty-seven cases did not show a gene copy number gain: a disomic pattern was found in 29/38 (76%) cases, and low polysomy was detected in 8/38 (21%). Only one case showed a copy number gain (high polysomy; figure 2).
A significant correlation was not demonstrated between the EGFR protein expression and gene amplification in MM.
Detection of EGFR mutations
Several different mutations located in different exons were found in MM. EGFR missense mutations were detected in six of the 38 patients (16%; table 4). Five tumours had missense mutations with amino acid substitutions within exon 20, resulting in the mutation of G to A single nucleotide polyporphism at codon 787 (glutamine to glutamine) in three tumours, mutation of C to T single nucleotide polyporphism at codon 785 (threonine to threonine) in one tumour, and mutation of asparagine to ricin at codon 816 (N816K) in one tumour. Another tumour had a mutation with amino acid substitutions within exon 18: threonine to methionine at codon 725 (T725M). One tumour had a mutation with amino acid substitutions within exon 21: glycine to glutamic acid at codon 857 (G857E). Only one case showed two incidences of mutation, within exon 20 and 21 (case no 28; figure 3). The difference in frequency of EGFR mutation between peritoneal and pleural MM (19% vs 14%) was negligible. Identical results of EGFR mutation were obtained using other DNA from another paraffin block committed to Mitsubishi Medicine. Corresponding background non-neoplastic regions of all the above MM harbouring mutation did not disclose any abnormalities.
We summarise the clinicopathological features, EGFR immunostaining scores and amplification for samples with EGFR mutations in table 5. No associations between EGFR mutation and other factors were confirmed.
MM is a rare malignancy with no known effective treatment. In the USA the incidence of MM has peaked as a result of the regulation of asbestos, but in Europe and Japan the incidence of MM will continue to increase because of the delay in legal regulation.1 2 A combination of exposure to environmental (asbestos) and infectious (SV 40) agents and somatic genetic alterations has been considered to be involved in the pathogenesis of MM, as recently reviewed elsewhere, but the molecular pathogenesis of MM is not well understood.17
EGFR expression in MM has been reported.9–11 18 19 Dazzi et al18 found EGFR expression (>5% cells) in 23 of 34 (68%) cases examined, with positive results more commonly found in the epithelial cell type of malignant pleural mesothelioma. EGFR expression was detected in 55% of malignant pleural mesothelioma patients in the study by Destro et al.19 The most recent reports about immunohistochemical EGFR detection in malignant pleural mesothelioma reported 46% and 32% positive immunostaining, respectively, which is probably related to the use of a more accurate semiquantitative scoring scale.9 10 Our figure of 53% positive immunohistochemical EGFR staining was similar to these results, and the increased frequency of EGFR in epithelioid MM was further confirmed.
Some reports indicate that EGFR overexpression is significantly more frequent in peritoneal than pleural MM (92% vs 33%, p=0.0004), but in our series no significant differences of frequency between pleural and peritoneal frequency were found.11
A previous report investigated the inﬂuence of immunohistochemical EGFR-positive status on survival. EGFR immunopositivity has in the past been indicated to be a poor prognostic factor in many solid tumours.20 To date, the influence of immunohistochemical EGFR-positive staining on the prognosis of MM is unclear. Some authors did not ﬁnd differences in survival when immunohistochemical EGFR-positive and negative staining were compared.10 19 Others suggested overexpression of the growth factor to be associated with a favourable outcome.18 21 Rena et al9 reported EGFR overexpression to be a factor negatively affecting prognosis in malignant pleural mesotheliomas. We will examine the association between EGFR expression and prognosis in the future.
We evaluated EGFR ampliﬁcation detected by FISH according to previously defined criteria for NSCLC. In our analysis, only one case with the epithelioid type of MM (3%) was shown to have the EGFR gene in high polysomy. A similar proportion was demonstrated by Okuda et al10 in 2008 in a study of 25 Japanese patients with malignant pleural mesothelioma; only one case of high polisomy (4%) was found by FISH and the patient had epithelial subtype malignant pleural mesothelioma. Therefore, EGFR protein overexpression in MM patients seems not to be related to gene copy number gain, as in other solid tumours.20
There have been several clinical trials of the EGFR TKI, gefitinib and erlotinib, in patients with malignant pleural mesothelioma.22 In the gefitinib trial, only two malignant pleural mesothelioma patients responded to gefitinib. These data suggest that a very small population (with EGFR gene amplification) may be candidates for anti-EGFR treatment. However, the results were less than satisfactory. Effective criteria for the optimal selection of patients who will benefit from these treatments are needed.
In this study, as a measure that may predict the efficacy of TKI, we looked for EGFR mutations in MM cases. In a previous report, EGFR mutations were found in 28.7% (27/94) of Japanese lung cancers, and the results of Taq-Man PCR were in complete agreement with the genomic DNA sequencing data.23 However, there were only a few incidences of EGFR mutation among our MM patients. Foster and colleagues24 25 reported that EGFR mutation was found in 31% of cases of malignant peritoneal mesothelioma. As far as we know, there has been no report of EGFR mutation of pleural MM in Japanese MM cases. In our study, EGFR missense mutations were detected in six out of 38 patients (16%) with pleural and peritoneal MM, one of which was found to have two mutation locations. Among the five mutation locations found in this study, two (N816K, G875E) were novel, two (T725M, Q787Q) were originally detected in NSCLC, and one (T785T) resembled a location previously noted for malignant peritoneal mesothelioma.24 Most of the EGFR mutations detected in this report showed rare patterns. Many researchers suspected that rare patterns of EGFR mutation detected by direct sequencing based on paraffin section were caused by PCR artifacts.26 We thus performed separate analyses using DNA from another paraffin block twice and committed them to Mitsubishi Medicine in mutation-positive cases. In this way, we confirmed that they were not caused by PCR artifacts.
In conclusion, our study confirmed that EGFR overexpression is a common feature in MM, especially the epithelial subtype. The protein overexpression of EGFR was not related to a gene copy number gain. To select candidates for anti-EGFR therapy among MM patients, further and larger studies will be needed.
We examined EGFR mutation in MM, compared with the result of immunohistochemical staining and FISH analysis.
EGFR missense mutations were detected in six of the 38 patients (16%) in MM.
EGFR mutation was not related to immunohistochemical and FISH analysis in MM.
The authors would like to thank Ms. Aya Shimada, Ms. Sawako Morikawa, Ms. Chika Ogawa, Ms.Tomoyo Adachi, and Mr. Takeshi Nishikawa for their technical assistance.
Competing interests None.
Patient consent Obtained.
Ethics approval Ethics approval was provided by the ethics committee of Nara Medical University.
Provenance and peer review Not commissioned; externally peer reviewed.