- Email: [email protected]
Detection of epidermal growth factor receptor and human epidermal growth factor receptor 2 activating mutations in lung adenocarcinoma by high-resolution melting amplicon analysis: correlation with gene copy number, protein expression, and hormone receptor expression
Human Pathology (2006) 37, 755 – 763
www.elsevier.com/locate/humpath
Detection of epidermal growth factor receptor and human epidermal growth factor receptor 2 activating mutations in lung adenocarcinoma by high-resolution melting amplicon analysis: correlation with gene copy number, protein expression, and hormone receptor expression Carlynn Willmore-Payne BS, MT (ASCP), Joseph A. Holden MD, PhD*, Lester J. Layfield MD Department of Pathology, University of Utah Health Sciences Center, Salt Lake City, UT 84108, USA Received 26 October 2005; revised 1 February 2006; accepted 6 February 2006
Keywords: Epidermal growth factor receptor activating mutation; HER2 activating mutation; Bronchioloalveolar carcinoma; High-resolution melting amplicon analysis
Summary Activating mutations in the epidermal growth factor receptor (EGFR) (7p12.3-p12.1) and in the human epidermal growth factor receptor 2 (HER2) (17q21.1) characterize a subset of lung carcinomas. These mutations may relate to the response of the tumor to the tyrosine kinase inhibitors gefitinib and erlotinib. High-resolution melting amplicon analysis is a screening technique that has been shown to be able to detect missense mutations as well as deletions and insertions in tumor DNA isolated from paraffin-embedded tissue sections. In this study, we used high-resolution melting amplicon analysis to screen for EGFR and human HER2 activating mutations in 39 patients with primary lung adenocarcinoma. There were 20 cases that showed bronchioloalveolar histology and 19 cases that did not. The EGFR exons screened were exons 18, 19, 20, and 21, and the HER2 exons screened were exons 19 and 20. Six (15%) of the 39 patients had tumors that contained EGFR activating mutations. Four of the mutations were in adenocarcinomas, which had some bronchioloalveolar features, and 2 mutations were in tumors without bronchioloalveolar features. The EGFR mutations were in exon 19 (2 cases), exon 20 (2 cases), and exon 21 (1 case). One case contained mutations in both exons 18 and 20. One (2.6%) of the 39 patients had a tumor that contained an HER2 activating mutation, and the mutation was located in exon 20. Two of the 6 EGFR mutation–positive cases showed polysomy for chromosome 7, and each one showed overexpression of EGFR as determined by immunohistochemical staining. The other EGFR mutation–positive cases did not show EGFR overexpression and appeared disomic for chromosome 7. The HER2 mutation–positive case was in an adenocarcinoma with bronchioloalveolar features. This tumor did not show overexpression of HER2 and was disomic for chromosome 17. For the non-EGFR mutation–positive cases, 4 (13%) of 32 evaluated cases showed polysomy for chromosome 7 and EGFR. No case showed EGFR gene amplification. Polysomy for chromosome 7 was not related to EGFR overexpression as estimated by immunohistochemistry.
4 Corresponding author. E-mail address: [email protected] (J. A. Holden). 0046-8177/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2006.02.004
756
C. Willmore-Payne et al. Estrogen and progesterone receptor expression was not strong in any of the cases and did not correlate with the presence of EGFR or HER2 mutations. D 2006 Elsevier Inc. All rights reserved.
1. Introduction Activated tyrosine kinases appear to be the causal event in a number of human malignancies [1]. The importance of this finding is reflected in the development of new anticancer drugs that specifically target these activated proteins. These new targeted drugs are less toxic than standard anticancer drugs and have shown remarkable success in the treatment of chronic myelogenous leukemia [2] and gastrointestinal stromal tumors (GISTs) [3,4]. Recently, a subset of patients with non–small cell lung cancer (NSCLC), who respond to the anticancer drugs gefitinib (Iressa, AstraZeneca Pharmaceuticals LP, Wilmington, Del) and erlotinib (Tarceva, OSI Pharmaceuticals Inc., Melville, NY), has been found. Gefitinib and erlotinib are small-molecule quinazolinamine derivatives, which target the active site of the epidermal growth factor receptor 1 (EGFR) [5]. Patients who respond to these drugs have been found to contain tumors with activating EGFR mutations [69]. The activated EGFR molecules allow for increased signaling of antiapoptotic pathways, which prevents tumor cell death [10]. Tumors with EGFR activation mutations become dependent upon this increased antiapoptotic pathway for survival, and when this pathway is inhibited by either gefitinib or erlotinib, the tumor cells die. Lung cancer cells with wild-type EGFR do not have this dependence on the antiapoptotic pathway for survival, which explains their resistance to gefitinib or erlotinib. It appeared from multiple clinical studies that the best predictor of response to gefitinib or erlotinib in patients with NSCLC is the presence of an underlying EGFR somatic mutation in the tumor [11]. These mutations are most common in women, in nonsmokers, in individuals of Asian descent, and in adenocarcinomas that show at least some bronchioloalveolar histology [12]. Mutations in EGFR make up about 10% of unselected cases of NSCLC in the United States but make up a larger proportion in Asian countries [5]. The activating mutations are clustered around the active site domain of EGFR and consist of in-frame deletions, insertions, or point mutations. The most common mutations include the deletion of 4 conserved amino acid residues (LREA) in exon 19 and a point mutation, L858R, in exon 21. Some patients with NSCLC may have, instead, activating mutations in the related protein, human epidermal growth factor receptor 2 (HER2), suggesting the possibility that these HER2 mutations might also sensitize the tumor to gefitinib or erlotinib [13,14]. For HER2, a point mutation in exon 19 (L755P) and in-frame insertions in exon 20 are observed. The initial optimism that the sole presence of an EGFR mutation in NSCLC might underlie the response of the
tumor to EGFR inhibitors has been tempered by several recent reports. These recent studies suggest that perhaps increased EGFR copy number or overexpression of the EGFR protein [15-17] may play a more important role in the sensitivity of NSCLC to EGFR inhibitors than the somatic EGFR mutations. Therefore, the role that EGFR or HER2 mutations play in the therapeutic response of lung cancer to gefitinib or erlotinib is not yet entirely clear. Perhaps there is a complicated interplay between EGFR or HER2 mutations with increased EGFR copy number or overexpression that determines the anticancer drug sensitivity or resistance in NSCLC. Nonetheless, determination of the presence or absence of EGFR or HER2 mutations in lung cancer might become an important part of the diagnostic workup of patients with NSCLC as well as an important part for future prospective clinical trials. Current technology to detect EGFR or HER2 activating mutations in lung cancer requires direct DNA sequencing [18]. Unfortunately, this is available only in large academic medical centers, and the results need to be interpreted by physicians experienced with DNA sequence analysis. We have recently reported on the use of high-resolution melting amplicon analysis (HRMAA) to detect c-kit and BRAF activating mutations in GISTs and melanomas [19,20]. In this report, we used HRMAA to detect EGFR and HER2 activating mutations in an unselected population of patients with lung adenocarcinoma. We found that HRMAA is able to detect EGFR or HER2 activating mutations from paraffin-embedded formalin-fixed lung cancer tissue. These results suggest that HRMAA might be used to provide molecular data for future clinical trials and perhaps help guide the treatment of patients with NSCLC.
2. Materials and methods 2.1. Sources of tissue The surgical pathology archive at the University of Utah Hospital, Salt Lake City, Utah, was searched for cases of adenocarcinoma of the lung from the years 1983 to 2005. Selection of cases was based on a sufficient amount of material available for mutation analysis. All histologic sections and immunohistochemical (IHC) stains were reviewed to confirm the diagnosis of a primary lung carcinoma. If not done, additional IHC stains were performed for cytokeratin 7 (CK 7), cytokeratin 20 (CK 20), and thyroid transcription factor-1 (TTF-1). Cases were subsequently classified as either adenocarcinoma with bronchioloalveolar (BAC) features or adenocarcinoma without bronchioloalveolar features in accord with recent recommendations [21]. For
EGFR and HER2 mutations in lung cancer the purposes of this study, cases of adenocarcinoma with bronchioloalveolar features included those tumors that were pure (noninvasive) BAC, BAC with focal invasion, and adenocarcinomas with focal BAC. A total of 39 cases were studied. Twenty were adenocarcinomas with BAC features, and 19 were adenocarcinomas without evidence of BAC histology. The use of human tissue for this study was approved by the Institutional Review Board at the University Of Utah Hospital (IRB 11903).
2.2. Antibodies and IHC staining Immunohistochemical staining for EGFR or HER2 was performed with either the DAKO (Carpinteria, Calif) pharmDX kit for EGFR or the DAKO Hercept test kit for HER2 in accord with the manufacturer’s instructions. Antibodies against CK 7 and CK 20 were also obtained from DAKO. Cytokeratin 7 was used at a dilution of 1:400, and CK 20 was used at a dilution of 1:200. Antibodies against TTF-1 were obtained from Novocastra (Newcastle upon Tyne, UK) and used at a dilution of 1:100. Antibodies against the estrogen receptor (ER) and the progesterone receptor (PR) were obtained prediluted from Ventana (Tucson, Ariz). Antigen retrieval techniques were used for all of the antibodies, and all IHC stains were performed with the Nexes instrument (Ventana Inc) in accord with the manufacturer’s instructions. The chromogen was diaminobenzidine. For the ER and PR immunostains, any staining was interpreted as positive without regard to intensity or percentage of cells positive. Interpretation of the EGFR or HER2 immunostains was done on a 0 to 3+ scale as described [22].
2.3. Chemicals, DNA, and enzymes These reagents have been described previously [19,20].
757 Table 1 EGFR exon
Epidermal growth factor receptor and HER2 primersa Sequence
Exon 18 (186 bp) Forward 5V-CTGAGGTGACCCTTGTCTCTGTGTTCTT-3V Reverse 5V-AGAGGCCTGTGCCAGGGACCTTA-3V Exon 19 (204 bp) Forward 5V-GCATGTGGCACCATCTCACAA-3V Reverse 5V-CCTGAGGTTCAGAGCCATGGA-3V Exon 20 (248 bp) Forward 5V-CCATGCGAAGCCACACTGA-3V Reverse 5V-CGTATCTCCCTTCCCTGATTACC-3V Exon 21 (236 bp) Forward 5V-GCAGAGCTTCTTCCCATGATGA-3V Reverse 5V-GCTGACCTAAAGCCACCTCCT-3V HER2 exon
Sequence
Exon 19 (214 bp) Forward 5V-CCACGCTCTTCTCACTCATATCC-3V Reverse 5V-AAGAGAGACCAGAGCCCAGAC-3V Exon 20 (252 bp) Forward 5V-GGGTGTGTGGTCTCCCATAC-3V Reverse 5V-GCAAAGAGCCCAGGTGCATA-3V a
Primers were designed as described in bMaterials and methodsQ section.
primer systems for their PCR and the amplicon size were confirmed by subjecting the PCR to polyacrylamide gel electrophoresis on nondenaturing 8.0% polyacrylamide gels as described [20].
2.4. Deoxyribonucleic acid isolation from paraffin-embedded tissue sections
2.6. Polymerase chain reaction and high-resolution melting curve analysis
These procedures have been previously described in detail [20]. Briefly, after reviewing the hematoxylin and eosin–stained slides, the corresponding paraffin block containing sufficient tumor tissue was selected. The area of tumor on the hematoxylin and eosin slide was outlined with a felt-tip pen, and the corresponding area in an unstained slide was microdissected with a scalpel and the microdissected tissue digested in a Tris buffer containing Tween 20 and proteinase K. After boiling for 10 minutes to inactivate the proteinase K, the digest was used directly for polymerase chain reaction (PCR).
These procedures have been described in detail [20]. Briefly, PCR was performed in a capillary cuvette (total volume, 20 lL) on a Light Cycler (Roche Diagnostics, Indianapolis, Ind). All reactions contained deoxyuridine triphosphate instead of deoxythymidine triphosphate. Before PCR, the samples were incubated with 1 unit of uracil Nglycosylase (Amperase; Applied Biosystems, Foster City, Calif) to prevent carryover contamination. All samples contained the double-stranded DNA binding dye, LCGreen (Idaho Technology, Salt Lake City, Utah). After an initial preincubation step at 958C (to denature the uracil glycosylase and activate the FastStart Taq DNA polymerase), the cycling conditions consisted of 45 cycles; each cycle included denaturation at 958C for 10 seconds, followed by an annealing at 608C for 10 seconds and, finally, extension at 748C for 0 seconds. Transition rates were 108C/s from denaturation to annealing, 18C/s from annealing to extension, and 208C/s from extension to denaturation. After PCR,
2.5. Design of primers Primers specific for EGFR exons 18, 19, 20, and 21 and for HER2 exons 19 and 20 were designed with the use of Primer Designer Software (Scientific and Education Software, Durham, NC). The sequences for the primers and the amplicon sizes are shown in Table 1. The integrity of the
758 Table 2
C. Willmore-Payne et al. Characteristics of EGFR and HER2 mutation–positive lung adenocarcinoma
EGFR mutation positive Case
Histologya
Mutation
1 2 3 4 5 6
BAC Non-BAC BAC BAC BAC Non-BAC
Exon Exon Exon Exon Exon Exon Exon
21 19 20 20 19 18 20
L858R del746-750 dup768-770 dup771-773 del746-752insV G719S S7681
EGFR IHCb
Cep 7c
EGFRc
Ratioc
Interpretation
3+ 0 Focal 3+ 0 0 0
5.35 1.73 4.00 ND 2.08 2.45
5.45 2.08 3.48 ND 2.08 2.35
1.02 1.2 0.86 ND 1 0.96
Polysomic Disomic Polysomic ND Disomic Disomic
HER2 mutation positive Case Histologya
Mutation
HER2 IHCb
Cep 17d
HER2d
Ratiod
Interpretation
7
Exon 20 G776Vins777C
0
1.78
1.93
1.08
Disomic
BAC
Abbreviation: ND, not determined. a Cases were classified as tumors with or without BAC features as described in bMaterials and methodsQ section. b Expression of EGFR or HER2 was estimated as described in bMaterials and methodsQ section and graded on a 0 to 3+ scale. c Fluorescence in situ hybridization for EGFR was performed as described in bMaterials and methodsQ section. The numbers represent the average gene copy number per cell, and the ratio is the copy number of EGFR divided by the copy number of chromosome 7. d Fluorescence in situ hybridization for HER2 was performed as described in bMaterials and methodsQ section. The numbers represent the average gene copy number per cell, and the ratio is the copy number of HER2 divided by the copy number of chromosome 17.
the samples were momentarily heated to 958C and then cooled to 408C. An appropriate Light Cycler cuvette containing the sample of interest was transferred to the High-Res Melter; a high-resolution DNA melting analysis instrument (Idaho Technology) and a melting analysis were performed as described [20]. Genomic DNA was used as a control. All samples were run in triplicate.
tion–positive cases contained either greater than 4 copies of the gene per cell or showed gene amplification (N2 copies of the gene per chromosome; HER2/CEP 17 N 2, EGFR/CEP 7 N 2). Fluorescence in situ hybridization–negative cases showed no amplification or had less than or equal to 4 copies of the gene per cell. These criteria have been defined previously [16].
2.7. Fluorescence in situ hybridization
2.8. Deoxyribonucleic acid sequence analysis
Fluorescence in situ hybridization (FISH), to detect EGFR and HER2 gene amplification, has been described previously [22,23]. For HER2 gene amplification, the PathVysion HER2 kit was used (Vysis, Downers Grove, Ill). For EGFR gene amplification, the LSI EGFR Spectrum Orange/CEP 7 Spectrum Green–labeled FISH probe was also obtained from Vysis. Briefly, unstained slides containing tumor tissue were deparaffinized, air dried, and protease treated, and the DNA was denatured. The slides were then washed successively with 70%, 95%, and 100% ethanol. After drying, the slides were mixed with hybridization solution containing the appropriate HER2 or EGFR probes. The slides were viewed under an Olympus BX 41 fluorescent microscope (Olympus, Tokyo, Japan) and interpreted by a single investigator (L. J. Layfield) who had passed the certification test supplied by the kit manufacturer. Cases were classified as either FISH positive or FISH negative. Fluorescence in situ hybridiza-
Bidirectional DNA sequencing was performed on all cases and done at the DNA sequencing core facility at the University of Utah. Analysis of the sequence was performed with the use of DNA Sequencher 4.1.4 software from Gene Codes Corporation, Ann Arbor, Mich.
3. Results 3.1. Patient characteristics Thirty-nine patients were identified with a diagnosis of adenocarcinoma of the lung. Immunohistochemical stains for CK 7, CK 20, and TTF-1 were performed on all the neoplasms. All cases evaluated, except one, showed strong CK 7 and TTF-1 immunostaining. Immunostaining for CK 20 was negative in most cases, although a few showed focal
Fig. 1 Histology of mutation-positive lung adenocarcinomas. The indicated photomicrographs show the histology of the mutation-positive lung adenocarcinomas (original magnification 250). The case numbers are the same as in Table 2. A, Case 1, adenocarcinoma with BAC features EGFR exon 21 mutation L858R. B, Case 2, adenocarcinoma without BAC features, EGFR exon 19 mutation, del746-750. C, Case 3, adenocarcinoma with BAC features, exon 20 mutation, dup768-770. D, Case 4, noninvasive BAC, EGFR exon 20 mutation, dup771-773. E, Case 5, noninvasive BAC, EGFR exon 19 mutation, del746-750insV. F, Case 6, adenocarcinoma without BAC features (papillary adenocarcinoma), EGFR mutations, exon 18 G719S and exon 20 S768I. G, Case 7, noninvasive BAC, HER2 mutation positive, exon 20, G776Vins777C.
EGFR and HER2 mutations in lung cancer
759
760 positive staining, but this was never as strong as CK 7 and TTF-1. One case was included in the study, which showed strong CK 7 and CK 20 positivity, but was negative for TTF1. This case was included because, histologically, it was clearly a BAC. Some BACs, especially the mucinous subtype, can show an immunophenotype of CK 7 positive, CK 20 positive, TTF-1 negative [28]. These IHC staining results support the diagnosis of a primary lung carcinoma in all cases studied. There were 17 males and 21 females, and in one case, the sex was not available. The ages ranged from 50 to 88 years with an average age of 66.5 F 8.7 years. All adenocarcinomas were further classified as adenocarcinoma with or without BAC features according to recent criteria [21]. There were 20 adenocarcinomas showing BAC features and 19 without. The adenocarcinomas with BAC features contained tumors that ranged from pure (noninvasive) BAC to invasive adenocarcinoma with only focal BAC features. The focal BAC histology in the invasive adenocarcinoma group was generally observed at the periphery of the tumor.
3.2. High-resolution amplicon melting analysis for EGFR exons 18, 19, 20, and 21 and for HER2 exons 19 and 20 All 39 cases of lung adenocarcinoma were screened by HRMAA to detect EGFR or HER2 activating mutations. Of the 39 total cases screened, 6 (15%) contained EGFR activating mutations and 1 (2.6%) contained an HER2 activating mutation. Mutations were more common in adenocarcinoma with BAC features. Of the 20 tumors with
C. Willmore-Payne et al. BAC features, 5 (25%) cases showed either EGFR or HER2 mutations, but in the 19 cases of tumors without BAC features, only 2 cases (11%) showed mutations. The characteristics of these mutation-positive cases are shown in Table 2. The corresponding histology of the 7 mutationpositive cases is shown in Fig. 1. Examples of the HRMAA curves obtained from the EGFR exon 21 point mutation (L858R) and the EGFR exon 19 deletion mutation (del746750) are shown in Fig. 2. A total of 234 exons were screened in this study by HRMAA (4 EGFR exons and 2 HER2 exons per case). To determine the accuracy of HRMAA, we followed up all melting curves by direct DNA sequencing. The melting curve of 1 (0.4%) of the 234 exons (HER2 exon 20) screened was indeterminate in that it could not be ascertained from visual inspection whether a mutation was present or not. Direct DNA sequence analysis indicated a normal sequence. One false negative was observed. In that case, the melting curve for EGFR exon 21 was interpreted as normal, but DNA sequencing indicated a mutation (L858R). Reevaluation of the histologic sections of that case showed numerous inflammatory cells, suggesting the mutation might have been obscured by normal cell contamination. Therefore, an additional block from the case was chosen with less inflammation, and care was taken to enrich in tumor cells during microdissection. Subsequent HRMAA for this case was now positive. This result emphasizes the importance of careful microdissection and optimal selection of the paraffin block. Titration experiments with mixtures of
Fig. 2 High-resolution melting amplicon analysis in lung adenocarcinoma. High-resolution melting amplicon analysis was performed as described in bMaterials and methodsQ section. A, High-resolution melting amplicon analysis shows an abnormal melting curve, and DNA sequence analysis indicated an EGFR exon 21 point mutation (L858R). The presence of a single base pair change is observed as an abnormal melting curve during HRMAA performed on DNA isolated from the tumor (dotted line). This tumor was estimated to contain about 50% normal cell contamination. The solid line represents normal genomic DNA. B, High-resolution melting amplicon analysis shows a marked abnormality in the melting curve obtained from DNA isolated from a lung adenocarcinoma (dotted line). Deoxyribonucleic acid sequence analysis indicated that the alteration was a deletion of amino acid residues 746 to 750 in the EGFR exon 19. The tumor was estimated to contain about 20% normal cell contamination. The solid line represents normal genomic DNA.
EGFR and HER2 mutations in lung cancer
761
3.3. Detection of silent mutations
Fig. 3 High-resolution melting amplicon analysis of EGFR exon 20. A common silent mutation is located in EGFR exon 20 at nucleotide position 2361, which results in a codon change from CAG to CAA. The silent mutation exists in the homozygous and heterozygous states and is easily detected by melting curve analysis. The melting curves of the homozygous wild type (CAG/CAG, solid line), the homozygous silent mutation (CAA/ CAA, short dotted line), and the heterozygous (CAG/CAA, long dotted line) are indicated.
wild-type and mutant DNA indicate that a single base pair (bp) change in amplicons 200 to 300 bp is easily detected when the base pair change makes up 25% or more of the total DNA (data not shown). Because most of the activating mutations are probably heterozygous, this suggests that microdissection on paraffin-embedded tissue should be performed on sections estimated to contain at least 50% tumor. In all other cases, a normal melting curve indicated a normal DNA sequence, and an abnormal melting curve indicated a mutation. No false-positive cases were observed.
A common silent mutation is located in EGFR exon 20 at nucleotide position 2361 (G to A). This results in a codon change from CAG to CAA, but it does not change the glutamine located at amino acid residue position 787. It has been described previously [9]. The high prevalence of this silent mutation complicates HRMAA for EGFR exon 20. As shown in Fig. 3, the homozygous wild type (CAG/CAG) has the highest melting temperature. The homozygous silent mutation (CAA/CAA) melts slightly below. As expected, the heterozygous silent mutation/wild type (CAA/CAG) melts with the lowest Tm. Therefore, cases in which the melting curve coincides with the homozygous wild type (showing the highest Tm) can be safely concluded to contain only the wild-type sequences without evidence of an activating mutation. All curves differing from this are ambiguous and require direct DNA sequencing. This also necessitates that the HRMAA control for EGFR exon 20 be the homozygous wild type. From our data, only 7 (18%) of 39 cases contained the homozygous wild type. This suggests that for screening purposes, only about 18% of cases will be able to be excluded from having an EGFR exon 20 activating mutation by HRMAA. In addition, a silent mutation was also discovered in exon 21. This is a change from C to T at nucleotide position 2508. We observed it in 1 (2.6%) of 39 cases. It does not change the arginine located at amino acid residue 836, but it must be distinguished from a true activating mutation in exon 21. Therefore, all abnormal melting curves for EGFR exon 21 require direct DNA sequencing. No other silent mutations were detected.
3.4. Estrogen and PR analysis Epidermal growth factor receptor and HER2 activating mutations are most common in women with NSCLC
Fig. 4 Expression of EGFR in mutation-positive lung adenocarcinoma. Expression of EGFR was determined as described in bMaterials and methodsQ section. The case numbers refer to the same cases as indicated in Table 2 and Fig. 1. A, Case 1, adenocarcinoma with the exon 21 L858R mutation. Fluorescence in situ hybridization indicated approximately 5 copies of chromosome 7 and EGFR per cell. The staining was interpreted as 3+. B, Case 3, adenocarcinoma with an exon 20 dup768-770. Fluorescence in situ hybridization indicated approximately 4 copies of chromosome 7 and EGFR per cell. The staining was interpreted as focal 3+.
762
C. Willmore-Payne et al.
[12,14]. To investigate whether this might relate to hormone receptor expression on the tumor cells, we performed estrogen and PR analysis on all cases by IHC. Because most cases were completely negative, any nuclear staining, regardless of intensity or the percentage of staining, was interpreted as positive. For the mutation-positive cases, 1 (14%) of 7 cases was interpreted as ER positive. None of the 7 mutation cases was interpreted as PR positive (data not shown). For the adenocarcinomas without mutation, 5 of the 32 cases were interpreted as ER positive (16%), and 8 of the adenocarcinomas without mutation were interpreted as PR positive (25%) (data not shown).
3.5. Epidermal growth factor receptor/HER expression and FISH To determine if the mutation-positive adenocarcinomas show EGFR or HER2 overexpression, we evaluated these tumors for overexpression by IHC. One of the 6 EGFR mutation–positive tumors showed uniform 3+ EGFR expression and another showed focal 3+ EGFR expression (Fig. 4). The other 4 were negative. Fluorescence in situ hybridization for EGFR indicated that both tumors with 3+ EGFR expression were polysomic for chromosome 7 and, therefore, contained additional copies of the EGFR gene, but the gene was not amplified. The negative EGFR IHC tumors were most likely disomic (Table 2). For the non-EGFR mutation cases (n = 33), FISH could be interpreted in 32 (97%). In these 32 cases, the average chromosome 7 number was 2.5 F 0.74 (range, 1.7-5.53) and the average EGFR gene copy number was also 2.5 F 0.95 (range, 1.6-5.45). Epidermal growth factor receptor expression as estimated by IHC was detectable in all 32 cases, but expression ranged from 1+ to 3+. There does not appear to be a correlation between EGFR expression determined by IHC, with EGFR FISH. Only 4 of the 32 nonmutation cases would be considered as FISH positive according to recent criteria [16]. This suggests that increased EGFR expression may occur independent of EGFR amplification. Possible mechanisms might include increased transcriptional activity of the EGFR gene or decreased degradation of the EGFR protein. Of these 4 FISH-positive Table 3 Correlation of EGFR expression with EGFR FISH in mutation-negative lung adenocarcinoma (n = 32) IHCa CEP 7b EGFRb Ratio FISH positive casesc
3+ (n = 11) 2.4 F 0.71 2.4 F 0.76 1 1
2+ (n = 14) 2.4 F 0.93 2.6 F 0.95 1.08 2
1+ (n = 7) 2.4 F 0.73 2.5 F 1.1 1.04 1
a Immunohistochemical was performed and interpreted as described in b Materials and methods Q section. b Fluorescence in situ hybridization with a LSI EGFR/CEP 7 probe as described in bMaterials and methodsQ section. The numbers represent the average gene copy number per cell. c Fluorescence in situ hybridization positive cases interpreted as described in bMaterials and methodsQ section.
cases, 3 (75%) had BAC features and the other one did not (25%). These results are summarized in Table 3. The HER2 mutation–positive tumor did not show overexpression of HER2 by IHC, and no amplification or polysomy of chromosome 17 was detected by FISH (Table 2). The possibility of HER2 overexpression in the other tumors was not addressed.
4. Discussion Gefitinib and erlotinib are active site tyrosine kinase inhibitors of EGFR. Increased EGFR expression has been observed in many human tumor systems, and therefore, drugs that inhibit EGFR would seem to be a rational choice of anticancer agents. Non–small cell lung cancer is one tumor that has consistently been reported to show increased EGFR expression [24]. However, it became clear from clinical studies that the expression of EGFR as determined by IHC did not relate to the response of patients with lung cancer to EGFR inhibitors [25]. Rather, it appeared that the response of patients with lung cancer to EGFR-targeted anticancer therapy was directly correlated with an underlying activating mutation in the EGFR gene occurring somatically in the tumor [6 -9]. Studies subsequent to this suggested that the response of NSCLC to EGFR inhibitors may be more complicated than this, and that EGFR gene copy number might be a better predictor of response than EGFR somatic mutation [15 -17]. These various studies emphasize that it is not entirely clear what parameter(s) in lung cancer might be the best predictor of response and overall survival to the EGFR inhibitors. This has lead to recent editorials suggesting the importance of future prospective clinical trials [26]. Because future clinical trials with EGFR inhibitors in the treatment of NSCLC will require mutation analysis, we investigated the potential of providing this information rapidly and efficiently with HRMAA. Unlike GIST or melanomas, where strong IHC expression serves to identify tumors with potential activating mutations for analysis, in NSCLC, all cases at the present time must be subjected to direct DNA sequencing to uncover potential mutations. With 4 EGFR exons and 2 HER2 exons to test, this can become overly burdensome. Unlike DNA sequencing, HRMAA can be accomplished in as little as 15 minutes after PCR of the appropriate exon. In the present study, as well as in previous work with HRMAA [19,27], all melting curves obtained by HRMAA that appeared normal (with one exception due to too much normal cell contamination) contained a normal sequence. We suggest, assuming careful microdissection to enrich in the component of tumor cells, that the presence of a normal melting curve indicates an underlying normal sequence, and direct DNA sequencing does not have to be performed. Because the frequency of EGFR mutations in NSCLC in the United States and Europe is only about 10% to 20%, HRMAA should be able to
EGFR and HER2 mutations in lung cancer eliminate the need for direct DNA sequencing in most cases for any future clinical trial involving American and European populations. Although an activating mutation can be inferred from an abnormal melting curve, we would advise that any abnormal melting curve still undergo direct DNA sequencing. This is necessary to uncover silent mutations as well as to give the exact type of activating mutation present. It is not known if all the EGFR or HER2 activating mutations are equally oncogenic or have the same drug inhibition constants. This information could be important for future clinical decisions. The presence or absence of an underlying EGFR or HER2 mutation in NSCLC may determine prognosis and therapy. High-resolution melting amplicon analysis may serve as a rapid screening procedure to yield this important information.
Acknowledgments We are grateful to the Associated and Regional Pathologists (ARUP) Institute for Clinical and Experimental Pathology, Salt Lake City, Utah, for support of this work. We thank Dr Carl Wittwer for his support of this project. We thank Trena Held for preparing the tables.
References [1] Dibb NJ, Dilworth SM, Mol CD. Switching on kinases: oncogenic activation of BRAF and the PDGFR family. Nature Rev 2004;4: 718 - 27. [2] Druker BJ, Talpaz M, Resta DJ, et al. Efficacy and safety of a specific inhibitor of the BCL-ABL tyrosine kinase in chronic myeloid leukemia. N Engl J Med 2001;344:1031 - 7. [3] Demetri GD, von Mehren M, Blanke CD, et al. Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med 2002;347:472 - 80. [4] Corless CL, Fletcher JA, Heinrich MC. Biology of gastrointestinal stromal tumors. J Clin Oncol 2004;22:3813 - 25. [5] Pao W, Miller VA. Epidermal growth factor receptor mutations, smallmolecule kinase inhibitors, and non–small-cell-lung cancer: current knowledge and future directions. J Clin Oncol 2005;23:2556 - 68. [6] 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 - 39. [7] Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004; 304:1497 - 500. [8] Pao W, Miller V, Zakowski M, et al. EGF receptor gene mutations are common in lung cancers from b never smokers Q and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A 2004;101:13306 - 11. [9] Takano T, Ohe Y, Sakamoto H, et al. Epidermal growth factor receptor gene mutations and increased copy numbers predict gefitinib sensitivity in patients with recurrent non–small-cell lung cancer. J Clin Oncol 2005;23:6829 - 37.
763 [10] Sordella R, Bell DW, Haber DA, et al. Gefitinib-sensitizing EGFR mutations in lung cancer activate anti-apoptotic pathways. Science 2004;305:1163 - 7. [11] Han SW, Kim TY, Hwang PG, et al. Predictive and prognostic impact of epidermal growth factor receptor mutation in non–small-cell lung cancer patients treated with gefitinib. J Clin Oncol 2005;23:2493 - 501. [12] 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 - 46. [13] Stephens P, Hunter C, Bignell G, et al. Lung cancer: intragenic ERBB2 kinase mutations in tumours. Nature 2004;431:525 - 6. [14] Shigematsu H, Takahashi T, Nomura M, et al. Somatic mutations of the HER2 kinase domain in lung adenocarcinomas. Cancer Res 2005;65:1642 - 6. [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 - 55. [16] Hirsch FR, Varella-Garcia M, McCoy J, et al. Increased epidermal growth factor receptor gene copy number detected by fluorescence in situ hybridization associates with increased sensitivity to gefitinib in patients with bronchioloalveolar carcinoma subtypes: a Southwest Oncology Group study. J Clin Oncol 2005;23:6838 - 45. [17] Tsao MS, Sakurada A, Cutz JC, et al. Erlotinib in lung cancer— molecular and clinical predictors of outcome. N Engl J Med 2005; 353:133 - 44. [18] Janne PA, Engelman JA, Johnson BE. Epidermal growth factor receptor mutations in non–small-cell lung cancer: implications for treatment and tumor biology. J Clin Oncol 2005;23:3227 - 34. [19] Willmore-Payne C, Holden JA, Tripp S, et al. Human malignant melanoma: detection of BRAF- and c-kit–activating mutations by high-resolution amplicon melting analysis. Hum Pathol 2005; 36:486 - 93. [20] Willmore-Payne C, Layfield LJ, Holden JA. c-KIT mutation analysis for diagnosis of gastrointestinal stromal tumors in fine needle aspiration specimens. Cancer 2005;105:165 - 70. [21] Travis WD, Garg K, Franklin WA, et al. Evolving concepts in the pathology and computed tomography imaging of lung adenocarcinoma and bronchioloalveolar carcinoma. J Clin Oncol 2005;23: 3279 - 87. [22] Tripp SR, Willmore-Payne C, Layfield LJ. Relationship between EGFR overexpression and gene amplification status in central nervous system gliomas. Anal Quant Cytol Histol 2005;27:71 - 8. [23] Willimore-Payne C, Holden JA, et al. Correlation of HER2 gene amplification with immunohistochemistry in breast cancer as determined by a novel monoplex polymerase chain reaction assay. Appl Immunohistochem Mol Morphol 2005;4:333 - 41. [24] Hirsch FR, Varella-Garcia M, Bunn Jr PA, et al. Epidermal growth factor receptor in non–small-cell lung carcinomas: correlation between gene copy number and protein expression and impact on prognosis. J Clin Oncol 2003;21:3798 - 807. [25] Parra HS, Cavina R, Latteri F, et al. Analysis of epidermal growth factor receptor expression as a predictive factor for response to gefitinib (dIressaT, ZD1839) in non–small-cell lung cancer. Br J Cancer 2004;91:208 - 12. [26] Johnson BE, Janne PA. Selecting patients for epidermal growth factor receptor inhibitor treatment: a FISH story or a tale of mutations? J Clin Oncol 2005;23:6813 - 6. [27] Willmore C, Holden JA, Zhou L, et al. Detection of c-kit–activating mutations in gastrointestinal stromal tumors by high resolution amplicon melting analysis. Am J Clin Pathol 2004;122:206 - 16. [28] Saad RS, Cho P, Silverman JF, et al. Usefulness of Cdx2 in separating mucinous bronchioloalveolar adenocarcinoma of the lung from metastatic mucinous colorectal adenocarcinoma. Am J Clin Pathol 2004;122:421 - 7.
www.elsevier.com/locate/humpath
Detection of epidermal growth factor receptor and human epidermal growth factor receptor 2 activating mutations in lung adenocarcinoma by high-resolution melting amplicon analysis: correlation with gene copy number, protein expression, and hormone receptor expression Carlynn Willmore-Payne BS, MT (ASCP), Joseph A. Holden MD, PhD*, Lester J. Layfield MD Department of Pathology, University of Utah Health Sciences Center, Salt Lake City, UT 84108, USA Received 26 October 2005; revised 1 February 2006; accepted 6 February 2006
Keywords: Epidermal growth factor receptor activating mutation; HER2 activating mutation; Bronchioloalveolar carcinoma; High-resolution melting amplicon analysis
Summary Activating mutations in the epidermal growth factor receptor (EGFR) (7p12.3-p12.1) and in the human epidermal growth factor receptor 2 (HER2) (17q21.1) characterize a subset of lung carcinomas. These mutations may relate to the response of the tumor to the tyrosine kinase inhibitors gefitinib and erlotinib. High-resolution melting amplicon analysis is a screening technique that has been shown to be able to detect missense mutations as well as deletions and insertions in tumor DNA isolated from paraffin-embedded tissue sections. In this study, we used high-resolution melting amplicon analysis to screen for EGFR and human HER2 activating mutations in 39 patients with primary lung adenocarcinoma. There were 20 cases that showed bronchioloalveolar histology and 19 cases that did not. The EGFR exons screened were exons 18, 19, 20, and 21, and the HER2 exons screened were exons 19 and 20. Six (15%) of the 39 patients had tumors that contained EGFR activating mutations. Four of the mutations were in adenocarcinomas, which had some bronchioloalveolar features, and 2 mutations were in tumors without bronchioloalveolar features. The EGFR mutations were in exon 19 (2 cases), exon 20 (2 cases), and exon 21 (1 case). One case contained mutations in both exons 18 and 20. One (2.6%) of the 39 patients had a tumor that contained an HER2 activating mutation, and the mutation was located in exon 20. Two of the 6 EGFR mutation–positive cases showed polysomy for chromosome 7, and each one showed overexpression of EGFR as determined by immunohistochemical staining. The other EGFR mutation–positive cases did not show EGFR overexpression and appeared disomic for chromosome 7. The HER2 mutation–positive case was in an adenocarcinoma with bronchioloalveolar features. This tumor did not show overexpression of HER2 and was disomic for chromosome 17. For the non-EGFR mutation–positive cases, 4 (13%) of 32 evaluated cases showed polysomy for chromosome 7 and EGFR. No case showed EGFR gene amplification. Polysomy for chromosome 7 was not related to EGFR overexpression as estimated by immunohistochemistry.
4 Corresponding author. E-mail address: [email protected] (J. A. Holden). 0046-8177/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.humpath.2006.02.004
756
C. Willmore-Payne et al. Estrogen and progesterone receptor expression was not strong in any of the cases and did not correlate with the presence of EGFR or HER2 mutations. D 2006 Elsevier Inc. All rights reserved.
1. Introduction Activated tyrosine kinases appear to be the causal event in a number of human malignancies [1]. The importance of this finding is reflected in the development of new anticancer drugs that specifically target these activated proteins. These new targeted drugs are less toxic than standard anticancer drugs and have shown remarkable success in the treatment of chronic myelogenous leukemia [2] and gastrointestinal stromal tumors (GISTs) [3,4]. Recently, a subset of patients with non–small cell lung cancer (NSCLC), who respond to the anticancer drugs gefitinib (Iressa, AstraZeneca Pharmaceuticals LP, Wilmington, Del) and erlotinib (Tarceva, OSI Pharmaceuticals Inc., Melville, NY), has been found. Gefitinib and erlotinib are small-molecule quinazolinamine derivatives, which target the active site of the epidermal growth factor receptor 1 (EGFR) [5]. Patients who respond to these drugs have been found to contain tumors with activating EGFR mutations [69]. The activated EGFR molecules allow for increased signaling of antiapoptotic pathways, which prevents tumor cell death [10]. Tumors with EGFR activation mutations become dependent upon this increased antiapoptotic pathway for survival, and when this pathway is inhibited by either gefitinib or erlotinib, the tumor cells die. Lung cancer cells with wild-type EGFR do not have this dependence on the antiapoptotic pathway for survival, which explains their resistance to gefitinib or erlotinib. It appeared from multiple clinical studies that the best predictor of response to gefitinib or erlotinib in patients with NSCLC is the presence of an underlying EGFR somatic mutation in the tumor [11]. These mutations are most common in women, in nonsmokers, in individuals of Asian descent, and in adenocarcinomas that show at least some bronchioloalveolar histology [12]. Mutations in EGFR make up about 10% of unselected cases of NSCLC in the United States but make up a larger proportion in Asian countries [5]. The activating mutations are clustered around the active site domain of EGFR and consist of in-frame deletions, insertions, or point mutations. The most common mutations include the deletion of 4 conserved amino acid residues (LREA) in exon 19 and a point mutation, L858R, in exon 21. Some patients with NSCLC may have, instead, activating mutations in the related protein, human epidermal growth factor receptor 2 (HER2), suggesting the possibility that these HER2 mutations might also sensitize the tumor to gefitinib or erlotinib [13,14]. For HER2, a point mutation in exon 19 (L755P) and in-frame insertions in exon 20 are observed. The initial optimism that the sole presence of an EGFR mutation in NSCLC might underlie the response of the
tumor to EGFR inhibitors has been tempered by several recent reports. These recent studies suggest that perhaps increased EGFR copy number or overexpression of the EGFR protein [15-17] may play a more important role in the sensitivity of NSCLC to EGFR inhibitors than the somatic EGFR mutations. Therefore, the role that EGFR or HER2 mutations play in the therapeutic response of lung cancer to gefitinib or erlotinib is not yet entirely clear. Perhaps there is a complicated interplay between EGFR or HER2 mutations with increased EGFR copy number or overexpression that determines the anticancer drug sensitivity or resistance in NSCLC. Nonetheless, determination of the presence or absence of EGFR or HER2 mutations in lung cancer might become an important part of the diagnostic workup of patients with NSCLC as well as an important part for future prospective clinical trials. Current technology to detect EGFR or HER2 activating mutations in lung cancer requires direct DNA sequencing [18]. Unfortunately, this is available only in large academic medical centers, and the results need to be interpreted by physicians experienced with DNA sequence analysis. We have recently reported on the use of high-resolution melting amplicon analysis (HRMAA) to detect c-kit and BRAF activating mutations in GISTs and melanomas [19,20]. In this report, we used HRMAA to detect EGFR and HER2 activating mutations in an unselected population of patients with lung adenocarcinoma. We found that HRMAA is able to detect EGFR or HER2 activating mutations from paraffin-embedded formalin-fixed lung cancer tissue. These results suggest that HRMAA might be used to provide molecular data for future clinical trials and perhaps help guide the treatment of patients with NSCLC.
2. Materials and methods 2.1. Sources of tissue The surgical pathology archive at the University of Utah Hospital, Salt Lake City, Utah, was searched for cases of adenocarcinoma of the lung from the years 1983 to 2005. Selection of cases was based on a sufficient amount of material available for mutation analysis. All histologic sections and immunohistochemical (IHC) stains were reviewed to confirm the diagnosis of a primary lung carcinoma. If not done, additional IHC stains were performed for cytokeratin 7 (CK 7), cytokeratin 20 (CK 20), and thyroid transcription factor-1 (TTF-1). Cases were subsequently classified as either adenocarcinoma with bronchioloalveolar (BAC) features or adenocarcinoma without bronchioloalveolar features in accord with recent recommendations [21]. For
EGFR and HER2 mutations in lung cancer the purposes of this study, cases of adenocarcinoma with bronchioloalveolar features included those tumors that were pure (noninvasive) BAC, BAC with focal invasion, and adenocarcinomas with focal BAC. A total of 39 cases were studied. Twenty were adenocarcinomas with BAC features, and 19 were adenocarcinomas without evidence of BAC histology. The use of human tissue for this study was approved by the Institutional Review Board at the University Of Utah Hospital (IRB 11903).
2.2. Antibodies and IHC staining Immunohistochemical staining for EGFR or HER2 was performed with either the DAKO (Carpinteria, Calif) pharmDX kit for EGFR or the DAKO Hercept test kit for HER2 in accord with the manufacturer’s instructions. Antibodies against CK 7 and CK 20 were also obtained from DAKO. Cytokeratin 7 was used at a dilution of 1:400, and CK 20 was used at a dilution of 1:200. Antibodies against TTF-1 were obtained from Novocastra (Newcastle upon Tyne, UK) and used at a dilution of 1:100. Antibodies against the estrogen receptor (ER) and the progesterone receptor (PR) were obtained prediluted from Ventana (Tucson, Ariz). Antigen retrieval techniques were used for all of the antibodies, and all IHC stains were performed with the Nexes instrument (Ventana Inc) in accord with the manufacturer’s instructions. The chromogen was diaminobenzidine. For the ER and PR immunostains, any staining was interpreted as positive without regard to intensity or percentage of cells positive. Interpretation of the EGFR or HER2 immunostains was done on a 0 to 3+ scale as described [22].
2.3. Chemicals, DNA, and enzymes These reagents have been described previously [19,20].
757 Table 1 EGFR exon
Epidermal growth factor receptor and HER2 primersa Sequence
Exon 18 (186 bp) Forward 5V-CTGAGGTGACCCTTGTCTCTGTGTTCTT-3V Reverse 5V-AGAGGCCTGTGCCAGGGACCTTA-3V Exon 19 (204 bp) Forward 5V-GCATGTGGCACCATCTCACAA-3V Reverse 5V-CCTGAGGTTCAGAGCCATGGA-3V Exon 20 (248 bp) Forward 5V-CCATGCGAAGCCACACTGA-3V Reverse 5V-CGTATCTCCCTTCCCTGATTACC-3V Exon 21 (236 bp) Forward 5V-GCAGAGCTTCTTCCCATGATGA-3V Reverse 5V-GCTGACCTAAAGCCACCTCCT-3V HER2 exon
Sequence
Exon 19 (214 bp) Forward 5V-CCACGCTCTTCTCACTCATATCC-3V Reverse 5V-AAGAGAGACCAGAGCCCAGAC-3V Exon 20 (252 bp) Forward 5V-GGGTGTGTGGTCTCCCATAC-3V Reverse 5V-GCAAAGAGCCCAGGTGCATA-3V a
Primers were designed as described in bMaterials and methodsQ section.
primer systems for their PCR and the amplicon size were confirmed by subjecting the PCR to polyacrylamide gel electrophoresis on nondenaturing 8.0% polyacrylamide gels as described [20].
2.4. Deoxyribonucleic acid isolation from paraffin-embedded tissue sections
2.6. Polymerase chain reaction and high-resolution melting curve analysis
These procedures have been previously described in detail [20]. Briefly, after reviewing the hematoxylin and eosin–stained slides, the corresponding paraffin block containing sufficient tumor tissue was selected. The area of tumor on the hematoxylin and eosin slide was outlined with a felt-tip pen, and the corresponding area in an unstained slide was microdissected with a scalpel and the microdissected tissue digested in a Tris buffer containing Tween 20 and proteinase K. After boiling for 10 minutes to inactivate the proteinase K, the digest was used directly for polymerase chain reaction (PCR).
These procedures have been described in detail [20]. Briefly, PCR was performed in a capillary cuvette (total volume, 20 lL) on a Light Cycler (Roche Diagnostics, Indianapolis, Ind). All reactions contained deoxyuridine triphosphate instead of deoxythymidine triphosphate. Before PCR, the samples were incubated with 1 unit of uracil Nglycosylase (Amperase; Applied Biosystems, Foster City, Calif) to prevent carryover contamination. All samples contained the double-stranded DNA binding dye, LCGreen (Idaho Technology, Salt Lake City, Utah). After an initial preincubation step at 958C (to denature the uracil glycosylase and activate the FastStart Taq DNA polymerase), the cycling conditions consisted of 45 cycles; each cycle included denaturation at 958C for 10 seconds, followed by an annealing at 608C for 10 seconds and, finally, extension at 748C for 0 seconds. Transition rates were 108C/s from denaturation to annealing, 18C/s from annealing to extension, and 208C/s from extension to denaturation. After PCR,
2.5. Design of primers Primers specific for EGFR exons 18, 19, 20, and 21 and for HER2 exons 19 and 20 were designed with the use of Primer Designer Software (Scientific and Education Software, Durham, NC). The sequences for the primers and the amplicon sizes are shown in Table 1. The integrity of the
758 Table 2
C. Willmore-Payne et al. Characteristics of EGFR and HER2 mutation–positive lung adenocarcinoma
EGFR mutation positive Case
Histologya
Mutation
1 2 3 4 5 6
BAC Non-BAC BAC BAC BAC Non-BAC
Exon Exon Exon Exon Exon Exon Exon
21 19 20 20 19 18 20
L858R del746-750 dup768-770 dup771-773 del746-752insV G719S S7681
EGFR IHCb
Cep 7c
EGFRc
Ratioc
Interpretation
3+ 0 Focal 3+ 0 0 0
5.35 1.73 4.00 ND 2.08 2.45
5.45 2.08 3.48 ND 2.08 2.35
1.02 1.2 0.86 ND 1 0.96
Polysomic Disomic Polysomic ND Disomic Disomic
HER2 mutation positive Case Histologya
Mutation
HER2 IHCb
Cep 17d
HER2d
Ratiod
Interpretation
7
Exon 20 G776Vins777C
0
1.78
1.93
1.08
Disomic
BAC
Abbreviation: ND, not determined. a Cases were classified as tumors with or without BAC features as described in bMaterials and methodsQ section. b Expression of EGFR or HER2 was estimated as described in bMaterials and methodsQ section and graded on a 0 to 3+ scale. c Fluorescence in situ hybridization for EGFR was performed as described in bMaterials and methodsQ section. The numbers represent the average gene copy number per cell, and the ratio is the copy number of EGFR divided by the copy number of chromosome 7. d Fluorescence in situ hybridization for HER2 was performed as described in bMaterials and methodsQ section. The numbers represent the average gene copy number per cell, and the ratio is the copy number of HER2 divided by the copy number of chromosome 17.
the samples were momentarily heated to 958C and then cooled to 408C. An appropriate Light Cycler cuvette containing the sample of interest was transferred to the High-Res Melter; a high-resolution DNA melting analysis instrument (Idaho Technology) and a melting analysis were performed as described [20]. Genomic DNA was used as a control. All samples were run in triplicate.
tion–positive cases contained either greater than 4 copies of the gene per cell or showed gene amplification (N2 copies of the gene per chromosome; HER2/CEP 17 N 2, EGFR/CEP 7 N 2). Fluorescence in situ hybridization–negative cases showed no amplification or had less than or equal to 4 copies of the gene per cell. These criteria have been defined previously [16].
2.7. Fluorescence in situ hybridization
2.8. Deoxyribonucleic acid sequence analysis
Fluorescence in situ hybridization (FISH), to detect EGFR and HER2 gene amplification, has been described previously [22,23]. For HER2 gene amplification, the PathVysion HER2 kit was used (Vysis, Downers Grove, Ill). For EGFR gene amplification, the LSI EGFR Spectrum Orange/CEP 7 Spectrum Green–labeled FISH probe was also obtained from Vysis. Briefly, unstained slides containing tumor tissue were deparaffinized, air dried, and protease treated, and the DNA was denatured. The slides were then washed successively with 70%, 95%, and 100% ethanol. After drying, the slides were mixed with hybridization solution containing the appropriate HER2 or EGFR probes. The slides were viewed under an Olympus BX 41 fluorescent microscope (Olympus, Tokyo, Japan) and interpreted by a single investigator (L. J. Layfield) who had passed the certification test supplied by the kit manufacturer. Cases were classified as either FISH positive or FISH negative. Fluorescence in situ hybridiza-
Bidirectional DNA sequencing was performed on all cases and done at the DNA sequencing core facility at the University of Utah. Analysis of the sequence was performed with the use of DNA Sequencher 4.1.4 software from Gene Codes Corporation, Ann Arbor, Mich.
3. Results 3.1. Patient characteristics Thirty-nine patients were identified with a diagnosis of adenocarcinoma of the lung. Immunohistochemical stains for CK 7, CK 20, and TTF-1 were performed on all the neoplasms. All cases evaluated, except one, showed strong CK 7 and TTF-1 immunostaining. Immunostaining for CK 20 was negative in most cases, although a few showed focal
Fig. 1 Histology of mutation-positive lung adenocarcinomas. The indicated photomicrographs show the histology of the mutation-positive lung adenocarcinomas (original magnification 250). The case numbers are the same as in Table 2. A, Case 1, adenocarcinoma with BAC features EGFR exon 21 mutation L858R. B, Case 2, adenocarcinoma without BAC features, EGFR exon 19 mutation, del746-750. C, Case 3, adenocarcinoma with BAC features, exon 20 mutation, dup768-770. D, Case 4, noninvasive BAC, EGFR exon 20 mutation, dup771-773. E, Case 5, noninvasive BAC, EGFR exon 19 mutation, del746-750insV. F, Case 6, adenocarcinoma without BAC features (papillary adenocarcinoma), EGFR mutations, exon 18 G719S and exon 20 S768I. G, Case 7, noninvasive BAC, HER2 mutation positive, exon 20, G776Vins777C.
EGFR and HER2 mutations in lung cancer
759
760 positive staining, but this was never as strong as CK 7 and TTF-1. One case was included in the study, which showed strong CK 7 and CK 20 positivity, but was negative for TTF1. This case was included because, histologically, it was clearly a BAC. Some BACs, especially the mucinous subtype, can show an immunophenotype of CK 7 positive, CK 20 positive, TTF-1 negative [28]. These IHC staining results support the diagnosis of a primary lung carcinoma in all cases studied. There were 17 males and 21 females, and in one case, the sex was not available. The ages ranged from 50 to 88 years with an average age of 66.5 F 8.7 years. All adenocarcinomas were further classified as adenocarcinoma with or without BAC features according to recent criteria [21]. There were 20 adenocarcinomas showing BAC features and 19 without. The adenocarcinomas with BAC features contained tumors that ranged from pure (noninvasive) BAC to invasive adenocarcinoma with only focal BAC features. The focal BAC histology in the invasive adenocarcinoma group was generally observed at the periphery of the tumor.
3.2. High-resolution amplicon melting analysis for EGFR exons 18, 19, 20, and 21 and for HER2 exons 19 and 20 All 39 cases of lung adenocarcinoma were screened by HRMAA to detect EGFR or HER2 activating mutations. Of the 39 total cases screened, 6 (15%) contained EGFR activating mutations and 1 (2.6%) contained an HER2 activating mutation. Mutations were more common in adenocarcinoma with BAC features. Of the 20 tumors with
C. Willmore-Payne et al. BAC features, 5 (25%) cases showed either EGFR or HER2 mutations, but in the 19 cases of tumors without BAC features, only 2 cases (11%) showed mutations. The characteristics of these mutation-positive cases are shown in Table 2. The corresponding histology of the 7 mutationpositive cases is shown in Fig. 1. Examples of the HRMAA curves obtained from the EGFR exon 21 point mutation (L858R) and the EGFR exon 19 deletion mutation (del746750) are shown in Fig. 2. A total of 234 exons were screened in this study by HRMAA (4 EGFR exons and 2 HER2 exons per case). To determine the accuracy of HRMAA, we followed up all melting curves by direct DNA sequencing. The melting curve of 1 (0.4%) of the 234 exons (HER2 exon 20) screened was indeterminate in that it could not be ascertained from visual inspection whether a mutation was present or not. Direct DNA sequence analysis indicated a normal sequence. One false negative was observed. In that case, the melting curve for EGFR exon 21 was interpreted as normal, but DNA sequencing indicated a mutation (L858R). Reevaluation of the histologic sections of that case showed numerous inflammatory cells, suggesting the mutation might have been obscured by normal cell contamination. Therefore, an additional block from the case was chosen with less inflammation, and care was taken to enrich in tumor cells during microdissection. Subsequent HRMAA for this case was now positive. This result emphasizes the importance of careful microdissection and optimal selection of the paraffin block. Titration experiments with mixtures of
Fig. 2 High-resolution melting amplicon analysis in lung adenocarcinoma. High-resolution melting amplicon analysis was performed as described in bMaterials and methodsQ section. A, High-resolution melting amplicon analysis shows an abnormal melting curve, and DNA sequence analysis indicated an EGFR exon 21 point mutation (L858R). The presence of a single base pair change is observed as an abnormal melting curve during HRMAA performed on DNA isolated from the tumor (dotted line). This tumor was estimated to contain about 50% normal cell contamination. The solid line represents normal genomic DNA. B, High-resolution melting amplicon analysis shows a marked abnormality in the melting curve obtained from DNA isolated from a lung adenocarcinoma (dotted line). Deoxyribonucleic acid sequence analysis indicated that the alteration was a deletion of amino acid residues 746 to 750 in the EGFR exon 19. The tumor was estimated to contain about 20% normal cell contamination. The solid line represents normal genomic DNA.
EGFR and HER2 mutations in lung cancer
761
3.3. Detection of silent mutations
Fig. 3 High-resolution melting amplicon analysis of EGFR exon 20. A common silent mutation is located in EGFR exon 20 at nucleotide position 2361, which results in a codon change from CAG to CAA. The silent mutation exists in the homozygous and heterozygous states and is easily detected by melting curve analysis. The melting curves of the homozygous wild type (CAG/CAG, solid line), the homozygous silent mutation (CAA/ CAA, short dotted line), and the heterozygous (CAG/CAA, long dotted line) are indicated.
wild-type and mutant DNA indicate that a single base pair (bp) change in amplicons 200 to 300 bp is easily detected when the base pair change makes up 25% or more of the total DNA (data not shown). Because most of the activating mutations are probably heterozygous, this suggests that microdissection on paraffin-embedded tissue should be performed on sections estimated to contain at least 50% tumor. In all other cases, a normal melting curve indicated a normal DNA sequence, and an abnormal melting curve indicated a mutation. No false-positive cases were observed.
A common silent mutation is located in EGFR exon 20 at nucleotide position 2361 (G to A). This results in a codon change from CAG to CAA, but it does not change the glutamine located at amino acid residue position 787. It has been described previously [9]. The high prevalence of this silent mutation complicates HRMAA for EGFR exon 20. As shown in Fig. 3, the homozygous wild type (CAG/CAG) has the highest melting temperature. The homozygous silent mutation (CAA/CAA) melts slightly below. As expected, the heterozygous silent mutation/wild type (CAA/CAG) melts with the lowest Tm. Therefore, cases in which the melting curve coincides with the homozygous wild type (showing the highest Tm) can be safely concluded to contain only the wild-type sequences without evidence of an activating mutation. All curves differing from this are ambiguous and require direct DNA sequencing. This also necessitates that the HRMAA control for EGFR exon 20 be the homozygous wild type. From our data, only 7 (18%) of 39 cases contained the homozygous wild type. This suggests that for screening purposes, only about 18% of cases will be able to be excluded from having an EGFR exon 20 activating mutation by HRMAA. In addition, a silent mutation was also discovered in exon 21. This is a change from C to T at nucleotide position 2508. We observed it in 1 (2.6%) of 39 cases. It does not change the arginine located at amino acid residue 836, but it must be distinguished from a true activating mutation in exon 21. Therefore, all abnormal melting curves for EGFR exon 21 require direct DNA sequencing. No other silent mutations were detected.
3.4. Estrogen and PR analysis Epidermal growth factor receptor and HER2 activating mutations are most common in women with NSCLC
Fig. 4 Expression of EGFR in mutation-positive lung adenocarcinoma. Expression of EGFR was determined as described in bMaterials and methodsQ section. The case numbers refer to the same cases as indicated in Table 2 and Fig. 1. A, Case 1, adenocarcinoma with the exon 21 L858R mutation. Fluorescence in situ hybridization indicated approximately 5 copies of chromosome 7 and EGFR per cell. The staining was interpreted as 3+. B, Case 3, adenocarcinoma with an exon 20 dup768-770. Fluorescence in situ hybridization indicated approximately 4 copies of chromosome 7 and EGFR per cell. The staining was interpreted as focal 3+.
762
C. Willmore-Payne et al.
[12,14]. To investigate whether this might relate to hormone receptor expression on the tumor cells, we performed estrogen and PR analysis on all cases by IHC. Because most cases were completely negative, any nuclear staining, regardless of intensity or the percentage of staining, was interpreted as positive. For the mutation-positive cases, 1 (14%) of 7 cases was interpreted as ER positive. None of the 7 mutation cases was interpreted as PR positive (data not shown). For the adenocarcinomas without mutation, 5 of the 32 cases were interpreted as ER positive (16%), and 8 of the adenocarcinomas without mutation were interpreted as PR positive (25%) (data not shown).
3.5. Epidermal growth factor receptor/HER expression and FISH To determine if the mutation-positive adenocarcinomas show EGFR or HER2 overexpression, we evaluated these tumors for overexpression by IHC. One of the 6 EGFR mutation–positive tumors showed uniform 3+ EGFR expression and another showed focal 3+ EGFR expression (Fig. 4). The other 4 were negative. Fluorescence in situ hybridization for EGFR indicated that both tumors with 3+ EGFR expression were polysomic for chromosome 7 and, therefore, contained additional copies of the EGFR gene, but the gene was not amplified. The negative EGFR IHC tumors were most likely disomic (Table 2). For the non-EGFR mutation cases (n = 33), FISH could be interpreted in 32 (97%). In these 32 cases, the average chromosome 7 number was 2.5 F 0.74 (range, 1.7-5.53) and the average EGFR gene copy number was also 2.5 F 0.95 (range, 1.6-5.45). Epidermal growth factor receptor expression as estimated by IHC was detectable in all 32 cases, but expression ranged from 1+ to 3+. There does not appear to be a correlation between EGFR expression determined by IHC, with EGFR FISH. Only 4 of the 32 nonmutation cases would be considered as FISH positive according to recent criteria [16]. This suggests that increased EGFR expression may occur independent of EGFR amplification. Possible mechanisms might include increased transcriptional activity of the EGFR gene or decreased degradation of the EGFR protein. Of these 4 FISH-positive Table 3 Correlation of EGFR expression with EGFR FISH in mutation-negative lung adenocarcinoma (n = 32) IHCa CEP 7b EGFRb Ratio FISH positive casesc
3+ (n = 11) 2.4 F 0.71 2.4 F 0.76 1 1
2+ (n = 14) 2.4 F 0.93 2.6 F 0.95 1.08 2
1+ (n = 7) 2.4 F 0.73 2.5 F 1.1 1.04 1
a Immunohistochemical was performed and interpreted as described in b Materials and methods Q section. b Fluorescence in situ hybridization with a LSI EGFR/CEP 7 probe as described in bMaterials and methodsQ section. The numbers represent the average gene copy number per cell. c Fluorescence in situ hybridization positive cases interpreted as described in bMaterials and methodsQ section.
cases, 3 (75%) had BAC features and the other one did not (25%). These results are summarized in Table 3. The HER2 mutation–positive tumor did not show overexpression of HER2 by IHC, and no amplification or polysomy of chromosome 17 was detected by FISH (Table 2). The possibility of HER2 overexpression in the other tumors was not addressed.
4. Discussion Gefitinib and erlotinib are active site tyrosine kinase inhibitors of EGFR. Increased EGFR expression has been observed in many human tumor systems, and therefore, drugs that inhibit EGFR would seem to be a rational choice of anticancer agents. Non–small cell lung cancer is one tumor that has consistently been reported to show increased EGFR expression [24]. However, it became clear from clinical studies that the expression of EGFR as determined by IHC did not relate to the response of patients with lung cancer to EGFR inhibitors [25]. Rather, it appeared that the response of patients with lung cancer to EGFR-targeted anticancer therapy was directly correlated with an underlying activating mutation in the EGFR gene occurring somatically in the tumor [6 -9]. Studies subsequent to this suggested that the response of NSCLC to EGFR inhibitors may be more complicated than this, and that EGFR gene copy number might be a better predictor of response than EGFR somatic mutation [15 -17]. These various studies emphasize that it is not entirely clear what parameter(s) in lung cancer might be the best predictor of response and overall survival to the EGFR inhibitors. This has lead to recent editorials suggesting the importance of future prospective clinical trials [26]. Because future clinical trials with EGFR inhibitors in the treatment of NSCLC will require mutation analysis, we investigated the potential of providing this information rapidly and efficiently with HRMAA. Unlike GIST or melanomas, where strong IHC expression serves to identify tumors with potential activating mutations for analysis, in NSCLC, all cases at the present time must be subjected to direct DNA sequencing to uncover potential mutations. With 4 EGFR exons and 2 HER2 exons to test, this can become overly burdensome. Unlike DNA sequencing, HRMAA can be accomplished in as little as 15 minutes after PCR of the appropriate exon. In the present study, as well as in previous work with HRMAA [19,27], all melting curves obtained by HRMAA that appeared normal (with one exception due to too much normal cell contamination) contained a normal sequence. We suggest, assuming careful microdissection to enrich in the component of tumor cells, that the presence of a normal melting curve indicates an underlying normal sequence, and direct DNA sequencing does not have to be performed. Because the frequency of EGFR mutations in NSCLC in the United States and Europe is only about 10% to 20%, HRMAA should be able to
EGFR and HER2 mutations in lung cancer eliminate the need for direct DNA sequencing in most cases for any future clinical trial involving American and European populations. Although an activating mutation can be inferred from an abnormal melting curve, we would advise that any abnormal melting curve still undergo direct DNA sequencing. This is necessary to uncover silent mutations as well as to give the exact type of activating mutation present. It is not known if all the EGFR or HER2 activating mutations are equally oncogenic or have the same drug inhibition constants. This information could be important for future clinical decisions. The presence or absence of an underlying EGFR or HER2 mutation in NSCLC may determine prognosis and therapy. High-resolution melting amplicon analysis may serve as a rapid screening procedure to yield this important information.
Acknowledgments We are grateful to the Associated and Regional Pathologists (ARUP) Institute for Clinical and Experimental Pathology, Salt Lake City, Utah, for support of this work. We thank Dr Carl Wittwer for his support of this project. We thank Trena Held for preparing the tables.
References [1] Dibb NJ, Dilworth SM, Mol CD. Switching on kinases: oncogenic activation of BRAF and the PDGFR family. Nature Rev 2004;4: 718 - 27. [2] Druker BJ, Talpaz M, Resta DJ, et al. Efficacy and safety of a specific inhibitor of the BCL-ABL tyrosine kinase in chronic myeloid leukemia. N Engl J Med 2001;344:1031 - 7. [3] Demetri GD, von Mehren M, Blanke CD, et al. Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med 2002;347:472 - 80. [4] Corless CL, Fletcher JA, Heinrich MC. Biology of gastrointestinal stromal tumors. J Clin Oncol 2004;22:3813 - 25. [5] Pao W, Miller VA. Epidermal growth factor receptor mutations, smallmolecule kinase inhibitors, and non–small-cell-lung cancer: current knowledge and future directions. J Clin Oncol 2005;23:2556 - 68. [6] 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 - 39. [7] Paez JG, Janne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 2004; 304:1497 - 500. [8] Pao W, Miller V, Zakowski M, et al. EGF receptor gene mutations are common in lung cancers from b never smokers Q and are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A 2004;101:13306 - 11. [9] Takano T, Ohe Y, Sakamoto H, et al. Epidermal growth factor receptor gene mutations and increased copy numbers predict gefitinib sensitivity in patients with recurrent non–small-cell lung cancer. J Clin Oncol 2005;23:6829 - 37.
763 [10] Sordella R, Bell DW, Haber DA, et al. Gefitinib-sensitizing EGFR mutations in lung cancer activate anti-apoptotic pathways. Science 2004;305:1163 - 7. [11] Han SW, Kim TY, Hwang PG, et al. Predictive and prognostic impact of epidermal growth factor receptor mutation in non–small-cell lung cancer patients treated with gefitinib. J Clin Oncol 2005;23:2493 - 501. [12] 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 - 46. [13] Stephens P, Hunter C, Bignell G, et al. Lung cancer: intragenic ERBB2 kinase mutations in tumours. Nature 2004;431:525 - 6. [14] Shigematsu H, Takahashi T, Nomura M, et al. Somatic mutations of the HER2 kinase domain in lung adenocarcinomas. Cancer Res 2005;65:1642 - 6. [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 - 55. [16] Hirsch FR, Varella-Garcia M, McCoy J, et al. Increased epidermal growth factor receptor gene copy number detected by fluorescence in situ hybridization associates with increased sensitivity to gefitinib in patients with bronchioloalveolar carcinoma subtypes: a Southwest Oncology Group study. J Clin Oncol 2005;23:6838 - 45. [17] Tsao MS, Sakurada A, Cutz JC, et al. Erlotinib in lung cancer— molecular and clinical predictors of outcome. N Engl J Med 2005; 353:133 - 44. [18] Janne PA, Engelman JA, Johnson BE. Epidermal growth factor receptor mutations in non–small-cell lung cancer: implications for treatment and tumor biology. J Clin Oncol 2005;23:3227 - 34. [19] Willmore-Payne C, Holden JA, Tripp S, et al. Human malignant melanoma: detection of BRAF- and c-kit–activating mutations by high-resolution amplicon melting analysis. Hum Pathol 2005; 36:486 - 93. [20] Willmore-Payne C, Layfield LJ, Holden JA. c-KIT mutation analysis for diagnosis of gastrointestinal stromal tumors in fine needle aspiration specimens. Cancer 2005;105:165 - 70. [21] Travis WD, Garg K, Franklin WA, et al. Evolving concepts in the pathology and computed tomography imaging of lung adenocarcinoma and bronchioloalveolar carcinoma. J Clin Oncol 2005;23: 3279 - 87. [22] Tripp SR, Willmore-Payne C, Layfield LJ. Relationship between EGFR overexpression and gene amplification status in central nervous system gliomas. Anal Quant Cytol Histol 2005;27:71 - 8. [23] Willimore-Payne C, Holden JA, et al. Correlation of HER2 gene amplification with immunohistochemistry in breast cancer as determined by a novel monoplex polymerase chain reaction assay. Appl Immunohistochem Mol Morphol 2005;4:333 - 41. [24] Hirsch FR, Varella-Garcia M, Bunn Jr PA, et al. Epidermal growth factor receptor in non–small-cell lung carcinomas: correlation between gene copy number and protein expression and impact on prognosis. J Clin Oncol 2003;21:3798 - 807. [25] Parra HS, Cavina R, Latteri F, et al. Analysis of epidermal growth factor receptor expression as a predictive factor for response to gefitinib (dIressaT, ZD1839) in non–small-cell lung cancer. Br J Cancer 2004;91:208 - 12. [26] Johnson BE, Janne PA. Selecting patients for epidermal growth factor receptor inhibitor treatment: a FISH story or a tale of mutations? J Clin Oncol 2005;23:6813 - 6. [27] Willmore C, Holden JA, Zhou L, et al. Detection of c-kit–activating mutations in gastrointestinal stromal tumors by high resolution amplicon melting analysis. Am J Clin Pathol 2004;122:206 - 16. [28] Saad RS, Cho P, Silverman JF, et al. Usefulness of Cdx2 in separating mucinous bronchioloalveolar adenocarcinoma of the lung from metastatic mucinous colorectal adenocarcinoma. Am J Clin Pathol 2004;122:421 - 7.