. Author manuscript; available in PMC: 2009 Oct 6.
Abstract IntroductionWe investigated EGFR and KRAS alterations among atypical adenomatous hyperplasia and small lung adenocarcinomas with bronchioloalveolar features to understand their role during multistage pathogenesis.
MethodsSixty lesions measuring 2 cm or less were studied, including 38 noninvasive lesions (4 atypical adenomatous hyperplasias, 19 Noguchi type A and 15 type B) and 22 invasive lesions (type C) based on the World Health Organization classification and Noguchi’s criteria. EGFR and KRAS mutations were examined using PCR-based assays. EGFR copy number was evaluated using fluorescence in situ hybridization.
ResultsEGFR and KRAS mutations were found in 26 (43.3%) and 5 (8.3%) lesions, respectively. Increased EGFR copy number status was identified in 10 lesions (16.7%), both mutant and wild type. EGFR or KRAS mutations were present in 39.5% and 7.9% (respectively) of noninvasive lesions and 50% or 9.1% (respectively) of invasive lesions. EGFR copy number was increased in 7.9% and 31.8% of noninvasive and invasive lesions (P = 0.029). Multivariate analysis revealed that increased EGFR copy number was the only significant factor to associate with invasive lesions (P = 0.035).
ConclusionsEGFR and KRAS mutations occur early during the multistage pathogenesis of peripheral lung adenocarcinomas. By contrast, increased EGFR copy number is a late event during tumor development and plays a role in the progression of lung adenocarcinoma independent of the initiating molecular events.
Keywords: Multistage pathogenesis, EGFR, KRAS, Mutation, Amplification
Adenocarcinomas of the lung represent a heterogeneous group of tumors. The major subtypes identified by the World Health Organization (WHO) consist of bronchioloalveolar carcinoma (BAC), acinar, papillary, or solid with mucin.1,2 However, the majority of tumors consist of mixtures of two or more subtypes. BAC is defined as a noninvasive tumor that has lepidic spread, i.e., replacement of the bronchiolar or alveolar epithelium by tumor cells. The WHO also recognized a peripheral lesion, atypical adenomatous hyperplasia (AAH), as a precursor lesion for BAC. These findings indicate a multistage progression model for most peripheral adenocarcinomas—AAHs giving rise to noninvasive BAC tumors, which later become invasive (mixed histology adenocarcinomas having a BAC component). In 1995, Noguchi and colleagues developed an independent classification for small peripheral adenocarcinomas (≤2 cm in size) consisting of six tumor subtypes.3 Types A, B, and C consisted entirely or predominantly of a BAC component (“replacement tumors”). Types A and B consisted of BAC tumors without (type A) or with (type B) areas of alveolar collapse and subsequent fibrosis. Among these two subtypes, nodal metastases were never present, and these tumors had a 100% 5-year survival if they were small (≤2 cm) and were completely resected. Type C tumors were BAC tumors with foci or fibroblastic proliferation. 28% of these tumors had lymph node metastases and the 5-year survival was 74.8%, similar to the average survival rate for small adenocarcinomas as a group. For practical purposes, we can regard the Noguchi type A and B tumors as the equivalent of WHO BAC subtype and Noguchi type C tumors as the equivalent of WHO mixed (i.e., invasive) tumors having a prominent BAC component. Certain molecular changes are characteristic of adenocarcinomas of the lung. These include mutations within the tyrosine kinase domain of the epidermal growth factor receptor (EGFR) gene, amplifications of the EGFR gene, and activating mutations of the KRAS gene. EGFR mutations, especially deletions in exon 19 and L858R point mutations in exon 21, target lung cancers arising in certain subsets: never smoker status, adenocarcinoma histology, female gender, and East Asian ethnicity.4–9 The mutations have been reported to be more frequent in adenocarcinomas having BAC features.4,6,10 The KRAS gene is a member of the ras family of oncogenes, and is located downstream of EGFR in the signaling pathway of the EGFR surface receptor. KRAS mutations are found in 10 to 30% in non-small cell lung cancer (NSCLC), especially adenocarcinomas, and over 90% of the mutations occurred in codon 12, occasionally in codon 13, and rarely in codon 61.11–15 Although both KRAS and EGFR mutations target lung adenocarcinomas, the former are associated with ever smoker status, male gender, and non-East Asian ethnicity.16,17 Thus, the two mutations seem to target different subpopulations and their presence may indicate alternative methods of adenocarcinoma pathogenesis.18 Of interest, EGFR and KRAS mutations may be present in AAH and in nonmalignant peripheral airways adjacent to cancers.19–21 Indeed, KRAS mutations or EGFR mutations in exon 19 or 21 cause multifocal BAC in mouse models.22,23 Increased copy number of EGFR, as assessed by fluorescence in situ hybridization (FISH) assay, was associated with lymph node metastasis, more advanced pathologic stage, and poor prognosis.24–28 This alteration has been described in NSCLC, especially adenocarcinomas, and may occur in mutant- or wild-type tumors.26,27 Furthermore, the three molecular changes discussed above may influence tumor response to small molecular weight tyrosine kinase inhibitors (TKIs) such as erlotinib and gefitinib. Both activating EGFR mutations and increased gene copy number are associated with good clinical outcome of EGFR-TKIs treatment,27,29,30 while KRAS mutations are associated with clinical resistance.30,31
To elucidate the roles of EGFR or KRAS mutations and increased EGFR copy number in the multistage pathogenesis of peripheral adenocarcinomas, we studied these molecular changes in small tumors having BAC features.
PATIENTS AND METHODS Patients and MaterialsDuring October 1997 to April 2005, 838 patients with NSCLC underwent pulmonary operations at Department of Cancer and Thoracic Surgery, Okayama University Hospital. Of these, 48 patients had lung adenocarcinoma (or AAH lesions) measuring 2 cm or less in greatest dimension and who had not received chemotherapy, radiotherapy, or targeted therapy before surgery. Among them, single lesions were resected from 39 patients. There were 8 patients who had 19 synchronous lesions in their resected specimens (2–4 lesions per resection) and one patient who had 2 metachronous lesions resected over a 6-year interval. Details of the synchronous and metachronous lesions, as defined by the criteria of Martini and Melamed,32 are presented in Table 1. Thus, from the 48 patients, a total of 60 lesions were available for study. Institutional review board permission and informed consent were obtained from all 48 patients for genetic analyses.
TABLE 1.Genetic Alternations in 21 Multiple Lesions of Nine Patients
Patient AgeThe following clinical factors were evaluated: age, sex, smoking status, and tumor size. All 60 cases were categorized into four histologic subtypes based on WHO classification and Noguchi’s criteria1,3; AAH (n = 4), type A (n = 19), type B (n = 15), and type C (n = 22). Because AAH is considered as a preinvasive lesion and Noguchi types A and B represent the noninvasive BAC (as defined by the WHO classification), we combined three categories and compared their findings with those present in Noguchi type C tumors (mixed subtype invasive adenocarcinomas with BAC component by the WHO classification). We excluded the other types of small adenocarcinomas (Noguchi types D, E, and F, WHO acinar, papillary, or solid subtypes) from this study as a convincing multistage pathogenesis has not been defined for these tumors.
DNA Extraction and Mutation Analyses of EGFR and KRAS GenesFor all 60 lesions that were formalin fixed and paraffin embedded, DNAs were isolated by DEXPAT (TaKaRa, Shiga, Japan) following the manufacturer’s instructions. Among them, DNAs of 17 frozen lesions were also isolated by digestion with proteinase K, followed by phenol-chloroform (1:1) extraction and ethanol precipitation from frozen specimen.33 EGFR mutations examination was limited to exon 19 deletions and exon 21 L858R point mutations by two methods: mutant-enriched or nonenriched PCR assays as described previously.34 Mutant-enriched PCR is a two-step PCR with intermittent restriction digestion to eliminate wild-type genes selectively, thus enriching the mutated genes at high sensitively. Nonenriched PCR is a one-step PCR without intermittent restriction digestion. The product of the amplification in the both methods was analyzed on 12% polyacrylamide gel electrophoresis (PAGE) via ethidium bromide staining. The common deletions of exon 19 were distinguished from the wild type based on PCR product length polymorphisms using 12% PAGE. For exon 21, Sau96I digestion, which could specifically digest the mutant type, was done before 12% PAGE. Mutant-enriched PCR assay can detect one mutant out of 2 × 103 wild-type genes. On the other hand, nonenriched PCR assay can detect one mutant of 10 wild-type genes. We analyzed KRAS mutations using a PCR assay with restriction digestion, followed by 12% PAGE, which was detectable of codon 12 point mutation based on the report by Kahn et al.35
These methods were predicted to detect about 85% of EGFR tyrosine kinase domain mutations8,36 and about 90% of KRAS mutations.11,12,37.
EGFR Copy Number Analysis by FISH AssayParaffin-embedded tissues of all 60 specimens were studied by FISH assay. A dual-color FISH assay was performed using the LSI EGFR spectrumOrange/CEP7 spectrumGreen probe (Vysis, Downers Grove, IL) following manufacturer’s instructions. At least 100 nonoverlapping inter phase nuclei per slide were scored by two independent observers (I.S., S.J.). All cases were classified into six FISH strata according to the criteria of Cappuzzo et al.27 Using their criteria, samples scored as having disomy, low trisomy, high trisomy, or low polysomy were classified as “FISH negative,” whereas samples scored as high polysomy or gene amplification were categorized as “FISH positive.”
Statistical AnalysesThe differences of significance among categorized groups were compared using χ2 or Fisher exact test tests as appropriate for univariate analyses. Logistic regression models using all variables were used to further explore observed differences and to identify baseline factors that may independently correlate with histologic classification. All data were analyzed with StatView 5.0 Program for Windows (SAS Institute Inc., Cary, NC). All statistical tests were two-sided and probability values <0.05 were defined as being statistically significant.
RESULTS Patient CharacteristicsThe details of patient characteristics are shown in Table 2. Pathologic examination of the 60 lesions from 48 patients indicated that each of the 60 lesions was a discreet presumably independently arising lesion. Of the 48 patients (9 of whom had multiple lesions), 31 had one or more noninvasive lesions whereas 17 had only invasive lesion. At the time of analysis, only one patient with an invasive cancer (which was EGFR FISH positive, but wild type for EGFR and KRAS mutations) had died of recurrent lung cancer 3 years after diagnosis. The mean follow-up duration of the 47 survivors were 35.0 months (range, 8.4–102.9 months). Thus, the relationship between survival and clinicopathologic or molecular features could not be analyzed.
TABLE 2.Characteristics of 60 Lesions in 48 Patients
Variables No. Age (yr) Median 65 (range, 41–84) <65 28 ≥65 32 Sex Male 23 Female 37 Smoking history Never 37 Ever 23 Tumor size (mm) Median 12 (range 4–20) ≤12 33 >12 27 Histology AAH 4 Type A 19 Type B 15 Type C 22Because some resected samples contained lesions of both noninvasive and invasive categories, we chose to present the data from the 60 individual lesions, rather than from 48 individual patients.
Somatic Alterations of EGFR and KRAS GenesDetails of genetic and clinical data are shown in Table 3 and Table 4. Among 60 lesions, we detected EGFR mutations in 26 lesions (43.3%), 15 as exon 19 deletions and 11 as L858R point mutation, with complete concordance of results by nonenriched and mutant-enriched PCR assays. KRAS codon 12 mutations were detected in 5 of 60 lesions (8.3%). Only one noninvasive cancer had mutations in both genes. We also confirmed the mutational data of EGFR and KRAS genes among 17 lesions whose DNAs were extracted from both paraffin-embedded samples and frozen samples. No discrepancies were identified from the analyses of these two differently handled DNA specimens.
TABLE 3.The Correlation Between Histologic Subsets and Clinical and Genetic Factors
Variables Subset TotalCorrelation Between Genetic Alteration and Clinical Factors
EGFR Mutation EGFR FISH KRAS Mutation Variables MutantFor EGFR copy number, 50 lesions (83.3%) were scored as FISH negative (disomy in 14 lesions, low trisomy in 19, high trisomy in 7, and low polysomy in 10), and 10 lesions (16.7%) were scored as FISH positive (high polysomy in 9 and amplification in 1). Four lesions had both EGFR mutation and FISH positivity and one lesion had both KRAS mutation and FISH positivity. The other 4 lesions as FISH positive were wild type for both genes. Thus, FISH positivity showed no selectivity for EGFR mutation positive (15.4%), KRAS mutation positive (20.0%), or wild type (16.7%) lesions. Example of FISH results are shown in Figure 1.
FIGURE 1.Representative examples of fluorescence in situ hybridization (FISH) with epidermal growth factor receptor (EGFR) (red signals) and chromosome 7 (green signals) probes. Case 1: One Noguchi type A lesion (left-top, H&E) with EGFR L858R point mutation (left-bottom) indicates EGFR FISH negative (Disomy) (right); Mean values of chromosome 7 and EGFR copy numbers are 2.19 ± 0.61 and 2.22 ± 0.69 (mean ± SD) per 100 nuclei, respectively. The percentage of ≤2 copies of EGFR gene is 90% and the ratio of EGFR gene to chromosome 7 is 1.00. Case 2: One Noguchi type C lesion (left-top, H&E) with EGFR exon 19 deletions (left-bottom) indicates EGFR FISH positive (high polysomy) (right); Mean values of chromosome 7 and EGFR copy numbers are 3.58 ± 1.30 and 3.36 ± 1.19 (mean ± SD) per 100 nuclei, respectively. The percentage of ≥4 copies of EGFR gene is 53% and the ratio of EGFR gene to chromosome 7 is 1.07.
Synchronous and Metachronous TumorsTwenty-one synchronous or metachronous lesions were found in nine patients (Table 1), consistent with the field cancer theory.32 One of these patients had metachronous lesions that arose in the opposite lungs 6 years apart (patient 1). We regarded these as independent primary tumors. Eight other patients had 2, 3, or 4 synchronous lesions (patients 2–9). Of these, 4 patients (patients 2–5) had individual lesions having either noninvasive or invasive components.
Evaluation of Clinical and Genetic Factors According to Histologic DifferenceWe investigated the relationship between clinical and genetic factors by each of the two histologic subsets (Table 3). By univariate analyses of the clinical factors, larger tumor size was significantly associated with invasive tumor (p = 0.027). Of the genetic factors, EGFR or KRAS mutations were present in 15 (39.5%) or 3 (7.9%) of 38 noninvasive lesions and 11 (50%) or 2 (9.1%) of 22 invasive lesions; however, these differences were not significant. By contrast, increased EGFR copy number was found in 3 (7.9%) of noninvasive lesions and 7 (31.8%) of invasive lesions (p = 0.029).
By multivariate analyses, FISH positivity was the only factor that was significantly associated with invasive tumors (OR = 6.5, 95% confidence interval: 1.14–37.3, p = 0.035).
DISCUSSIONOur study reveals important differences in the sequential appearance of molecular changes during multistage pathogenesis of small peripheral adenocarcinomas having a BAC component. EGFR mutations were found in all four categories of lesions examined including an AAH. KRAS mutations, while not detected in the small number of AAH lesions examined, were present in noninvasive and invasive carcinomas of Noguchi types A, B, and C. There were no significant differences in the frequencies of these two mutation types between the pathologic types. By contrast, we found a significant difference for EGFR gene FISH status, with FISH positivity being significantly more frequent in the invasive group of lesions (type C).
Our findings regarding these mutational changes are consistent with the published literature. Both KRAS and EGFR mutations have been described in AAH lesions,20,37,38 and EGFR mutations have been found in the nonmalignant peripheral airways in the vicinity of invasive peripheral adenocarcinomas,21 indicating that mutations of both genes are early events that play a role in tumor initiation.
Recent reports demonstrate that the KRAS mutations are more frequent in BACs of the mucinous subtype whereas EGFR mutations are more frequent in the nonmucinous subtype.10,39 In our 60 lesions, 59 lesions (98.3%) were nonmucinous type of BACs, and only one lesion was a mucinous BAC without genetic alterations. Furthermore, we focused on small adenocarcinomas with BAC components, which are well-known to correlate with female, never smoking status.40 KRAS mutations are also associated with ever smoker status, male gender, and non-East Asian ethnicity.16,17 These findings support the lower frequency of KRAS mutations in this cohort.
We found one rare case harboring both EGFR and KRAS mutations, who is a 61-year-old man with heavy smoking history (50 packs per year). This noninvasive and nonmucinous lesion (Noguchi type B) has KRAS codon 12 point mutation and EGFR L858R point mutation without EGFR FISH positivity (low trisomy). As mentioned above, KRAS mutations are significantly frequent in males and ever smokers whereas EGFR mutations in females and never smokers and BACs, especially in the nonmucinous type.10,16,17,39 However, Shigematsu et al. also indicated that EGFR mutations could be detected in 10% of males and 14% of ever smokers.16 These findings might explain this rare case, which has both characteristics of EGFR and KRAS mutations.
Our current study clearly revealed that increased copy number of the EGFR gene seems to be a progression event independent from the initial alterations, being more than four times more frequent in invasive lesions than in noninvasive lesions. These differences are also significant even if the four AAH lesions are removed from the analysis (p = 0.039). Patient number 4 (Table 1) is of particular interest. She had synchronous noninvasive AAH and invasive (Noguchi type C) lesions. Although both lesions had identical (L858R) EGFR mutations, the invasive cancer was FISH positive, whereas the AAH was FISH negative, providing support for our concept that the activating mutations occur early in pathogenesis, while FISH positivity is a late event associated with advanced (invasive) cancers. Our findings are consistent with recent reports that describe EGFR mutations preceding amplification, with amplification favoring the mutant allele.41 Our findings clearly demonstrate that FISH positivity occurs as frequently in mutant- and wild-type tumors, and that these two events associated with response to tyrosine kinase inhibitors occur independently of each other.26,42
Gazdar et al. have proposed that peripheral lung adenocarcinomas arise via three or more pathways.18 In smokers, activation of the KRAS pathway via mutations occurs, whereas in nonsmokers, there is upstream activation of the EGFR pathway via activating mutations. As these changes only account for about 50% of lung adenocarcinomas, other initiating events may also contribute to pathogenesis. Based on our finding, a hypothesis of putative molecular pathways to peripheral lung adenocarcinomas is shown in Figure 2. Although EGFR and KRAS mutations are early, possibly initiating events for subsets of adenocarcinomas, increased copy number of the EGFR gene seems to be a late or progression event and targets adenocarcinomas irrespective of the initiating events. However, our sample size is not large and further investigation is warranted to confirm our hypothesis.
FIGURE 2.Putative multiple molecular pathways to peripheral lung adenocarcinomas. Adenocarcinomas represent a heterogenous mixture, and may arise from peripheral or central airways. Adenocarcinomas arising from the peripheral airways often contain a bronchioloalveolar component. From work presented in this report and from the published literature, it would appear that there are at least three early or initiating sets of events. Pathway 1, more frequently seen in never or light smokers, involves activating mutations of the epidermal growth factor receptor (EGFR) gene. Pathway 3, almost limited to ever smokers, involves activating mutations of the KRAS gene. Pathway 1 targets females and East Asian ethnicity, whereas pathway 3 targets males and non-East Asian ethnicity. However, the combined frequencies of these mutations (on a worldwide basis) are about 50% of adenocarcinomas. Thus, one or more alternative pathways (pathway 2) must use other initiating events. From data presented in this report, it would appear that increased EGFR copy number is a late (progression) event, and that targets tumors irrespective of the initiating events. Figure modified from Gazdar AF, Shigematsu H, Herz J, et al. Mutations and addiction to EGFR: the Achilles ‘heal’ of lung cancers? Trends Mol Med 2004;10:481–486; and Wistuba II, Gazdar AF. Lung cancer preneoplasia. Annu Rev Pathol Mech Dis 2006;1:331–348.
Our findings suggest that increased gene copy number may be a relatively late event in tumor pathogenesis, and that the original molecular status of the tumor may not accurately reflect the response to TKIs of a recurrent or metastatic lesion. The recognition that molecular events are dynamic and progressive necessitates retesting advanced or recurrent tumors to determine optimal, individualized targeted therapies.
ACKNOWLEDGMENTSThe authors thank Mrs. Makiko Tabata, Cancer and Thoracic surgery, Hiroyuki Kugoh, MD, and Kazuhiro Murakami, MD, Division of Molecular and Cell Genetics, Molecular and Cellular biology, Faculty of Medicine, Tottori University for technical support of FISH assay. The authors thank Xian-Jin Xie, PhD, Harold C. Simmons Comprehensive Cancer Center, University of Texas Southwestern Medical Center, Dallas, for support of statistical analyses. Masayuki Noguchi, MD, Department of Pathology, Institute Basic Medical Sciences, University of Tsukuba, kindly provided information regarding pathologic classification.
Supported by a Grant-in Aid for Scientific Research from the Ministry of Education, Science, Sports, Culture and Technology of Japan (18790993 for S.T.) and a grant from the Specialized Program of Research Excellence in Lung Cancer (P50CA70907) from the National Cancer Institute, Bethesda, Maryland, USA.
FootnotesDisclosure: The authors declare no conflict of interest.
REFERENCESRetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.3
