Abstract
Aim: We performed this study to evaluate the frequency of epidermal growth factor receptor (EGFR) mutations and their association with the histological subtype of lung adenocarcinoma diagnosed via small biopsy specimens. Patients and Methods: Three hundred and fifty-nine lung adenocarcinoma specimens were tested for EGFR mutation by a direct sequencing method. In 135 patients, histological subtypes were classified according to the 2011 International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society classification. We compared the EGFR mutation frequency by clinicopathological characteristics. Results: We detected 135 (37.6%) EGFR mutations and the incidence was highest in women who never smoked (54.6%). With regard to histological subtype, the highest prevalence of EGFR mutation was found in papillary (81.3%), followed by lepidic (70.4%), acinar (58.1%) and solid (28.3%) tumor types. In addition, the expression of thyroid transcription factor-1 was significantly higher in tumors with the EGFR mutation than in those without (p<0.001). Conclusion: The histological subtype of adenocarcinoma can be predictive of existing EGFR mutation, although the histology was confirmed using only small biopsies.
The most important recent advance in the therapy of patients with non-small cell lung cancer (NSCLC) was the discovery of the epidermal growth factor receptor (EGFR) gene, which has explained why some patients have an excellent response to an oral tyrosine kinase inhibitor (TKI) of this receptor (1, 2). In particular, EGFR-TKIs as a first-line treatment are recommended for patients with NSCLC who harbor an EGFR mutation, replacing conventional cytotoxic chemotherapy, due to good and rapid response (3-5). Therefore, EGFR mutation analysis is necessary and important in clinical practice.
Most patients with advanced lung cancer are diagnosed only through small biopsies, such as those obtained by percutaneous fine-needle aspiration or bronchoscopic biopsy. These patients do not have the opportunity to undergo curative surgery. Therefore, additional tests such as immunohistochemical (IHC) staining and mutation analysis are carried out from these small biopsy samples. Sometimes, adequate specimens for additional analysis cannot be obtained because of the patient's condition or the location of the cancer. The specimens obtained through these procedures are also often inadequate for DNA extraction for EGFR mutation analysis due to sample degradation or lack of specimen volume. To optimize the treatment of these patients, it is helpful to investigate suitable surrogate markers for predicting the presence of EGFR mutations.
There have been many reports on the association of EGFR mutation with the clinicopathological features of NSCLC, but most of them have been studied based on resection specimens (6, 7). However, the status of EGFR mutations in small biopsy specimens is still unclear. This study aimed to investigate the relationship between EGFR mutations and the clinicopathological features of lung adenocarcinomas as diagnosed with small biopsy specimens.
Patients and Methods
Patients and specimens. All clinicopathological information was obtained from the medical records and pathology reports at the Yeungnam University Medical Center, Daegu, Korea. Three hundred and fifty-nine patients with lung adenocarcinoma (356 biopsies and 3 resections) from January 2008 to December 2011 were included in this study. Among 356 biopsies, 282 patients were diagnosed by percutaneous needle biopsy, 74 by bronchoscopic biopsy and three by resection. Patients eligible for this study had to meet the following criteria: pathologically-confirmed adenocarcinoma of primary in lung and known status for EGFR mutations. Patients who previously received chemotherapy or radiotherapy were excluded. This study was approved by the Institutional Review Board of the Yeungnam University Medical Center (Approval number YUH-12-0358-O29). Written informed consent for participation was obtained from each patient.
Clinical and pathological variables. Clinical and pathological data collected for analysis included sex, age at diagnosis, cumulative smoking dose before diagnosis, clinical TNM stage and histological subtype. Light smokers were defined as patients who had a total of 10 pack-years or more of smoking, and heavy smokers were defined as those who had a total of less than10 pack-years of smoking. Former smokers were defined as those who had stopped smoking at least 10 years previously. The clinical stage was determined based on bronchoscopy, chest computed tomographic (CT) scan, brain magnetic resonance imaging (MRI), bone scan and torso positron-emission tomography (PET) using the seventh edition of the American Joint Committee on Caner (AJCC) criteria (8). The histological subtypes of the specimens were classified according to the 2004 WHO classification (224 patients) (9) and the new International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society (IASLC/ATS/ERS) multidisciplinary classification (135 patients) (10). EGFR mutation analysis according to histological subtype based on 2011 IASLC/ATS/ERS classification was carried out in 135 patients.
Analysis of EGFR mutation. All the specimens were investigated for EGFR mutation, exon 18-21 of the tyrosine kinase domain, by polymerase chain reaction (PCR) and the direct DNA sequencing method (2). Genomic DNA was extracted from paraffin-embedded tissues. After de-paraffinization with xylene, tissue sections were stained with hematoxylin and eosin, and target lesions were selectively obtained to minimize the presence of any normal tissue. Details about the EGFR mutation types and the methodologies for mutation detection have been published elsewhere (11). It was investigated whether lung adenocarcinoma belonging to a certain pathological subtype was related to EGFR mutation or not.
Thyroid transcription factor 1 (TTF1) expression. Three hundred and forty-eight specimens were evaluated for the expression of TTF1 protein by IHC staining. A positive case was defined as the presence of nuclear TTF1 expression.
Statistical analysis. Statistical analysis for categorical variables was performed using chi-square test or Fisher's exact test when an expected value was less than five. Independent samples t-test was used to test for differences between EGFR mutation-positive and-negative groups. All the statistical analyses were conducted with SPSS software for Windows, version 18.0 (SPSS Inc., Chicago, IL, USA). Statistical significance was determined by a two-tailed p-value less than 0.05.
Results
Patients' characteristics. A total of 359 patients with lung adenocarcinoma (217 male patients, 142 female patients) were eligible for the study. All patients were Korean. The clinical characteristics of the patients are described in Table I. The median age at diagnosis was 68.0 (range 31-89) years old. One hundred and fifty-three patients were never-smokers and 203 patients were current or ever-smokers. The smoking history of three patients was not clear. The majority of patients had disease of clinical TNM stage III or IV (Table I).
Pathological features according to the 2011 IASLC/ATS/ERS classification. Among 135 patients, classified on the basis of the 2011 IASLC/ATS/ERS classification system, tumor in 101 was of the pure type (74.8%) and in 34 was of mixed type (25.6%). For the 101 with pure-type adenocarcinoma, the most common histological pattern was acinar (44.6%), and among 34 patients with mixed-type adenocarcinoma, the most common main histology was also acinar (41.2%, 14 out of 34) (Table II). When classifying all the patients based on the main histological component, the most common sub-type was acinar (43.7%, 59/135), followed by solid (26.7%, 36/135), lepidic (13.3%, 18/135), papillary (6.7%, 9/135), mucinous (5.2%, 7/135) and micropapillary adenocarcinoma (4.4%, 6/135).
EGFR mutation and clinical features. EGFR mutations were present in 37.6% (135/359) of lung adenocarcinomas. As shown in Table I, patients with EGFR mutations were significantly younger than EGFR wild-type patients (mean age, 63.9 vs. 66.5 years; p=0.035). Mutations were found more frequently in women than in men (51.4% vs. 28.6%, p<0.001), and in never-smokers than in smokers (54.2% vs. 25.6%, p<0.001).
We also evaluated the relationship between EGFR mutation and the detailed status of smoking history. There was no difference between former smokers and current smokers. However, the incidence of EGFR mutations in light smokers was significantly higher than in heavy smokers (55.2% vs. 20.7%, p<0.001).
In men and ever-smokers, there was a significant number of EGFR mutations. The frequency of EGFR mutations in men with adenocarcinoma who smoked was 26.0% (50/192). Moreover, the number of patients with EGFR mutations were significantly higher in men who never smoked or smoked lightly (52.2% and 62.5%) compared to heavy smokers (20.7%).
The current study population distribution according to the 7th edition of the AJCC (8) cancer staging system is shown in Table I. More than half of the cases already had locally or systemically advanced lung cancer at diagnosis, and the incidence of EGFR mutation in these differed between the clinical stages. However, the incidence of EGFR mutations was not different according to the diagnostic method used (Table I).
EGFR mutation and pathological features. The frequency of EGFR mutations was higher in those with mixed-subtype than in those with pure-subtype tumors (64.7% vs. 37.6%) (Table II). The distribution of EGFR mutations according to main histological subtype is shown in Table III. EGFR mutations occurred at a higher frequency in patients with papillary, lepidic and acinar pattern (77.8%, 61.1%, and 57.6%). But its frequency was low in those with solid histological pattern (16.7%).
On the basis of the findings of histology, we further classified patients according to specific components. Specimens that contained acinar component accounted for 54.8% of the total, followed by solid (34.1%), lepidic (20.0%) and papillary (11.9%) components. EGFR mutations were more frequently observed in lung adenocarcinomas with acinar, lepidic and papillary components, and conversely, infrequently observed in those with solid and mucinous components (Table IV). All 10 cases with a mucinous adenocarcinoma component and 1 case with signet ring cell had a negative EGFR mutation status. In addition, most of the EGFR-mutant adenocarcinomas (95.5%) were also TTF1 positive (Table V).
EGFR mutation profile. A total of 359 adenocarcinomas were examined for EGFR mutations in exons 18 to 21 of the EGFR gene. The majority of mutations were exon 19 deletions (60.7%, 82 out of 135). Forty mutations were found at exon 21, 12 at exon 20 and six at exon 18. Seven patients had mutations involving multiple sites and one of them had triple mutations. Known as the TKI-resistant mutation, T790M was found in two patients who were naïve to TKI. However, both of them also had EGFR-sensitive mutations (Table VI).
Discussion
This investigation showed that despite tumor heterogeneity, the histological characteristics of small biopsies can predict the presence of EGFR mutation. Based on the research that EGFR mutation-positive patients with extremely poor performance status (ECOG 3 to 4) can benefit from first-line EGFR-TKI, this result could be useful in clinical practice to optimize the management of patients who are diagnosed with NSCLC at an advanced stage with poor general condition (3).
In clinical practice, EGFR mutational status using small biopsy specimens is very important because most of patients are diagnosed with lung cancer at a late stage. In lung cancer, the results of molecular studies using small biopsy specimens are underestimated because of the hypothesis of intratumoral heterogeneity. Some studies have postulated the heterogeneous distribution of EGFR mutation within individual tumors (12, 13). However, recent comprehensive studies suggest that heterogeneous distribution of EGFR mutations is very rare (14, 15).
In this study, the incidence of activating EGFR mutations in lung adenocarcinomas was 37.6% (135 outof 359) which is much lower than that found in other reports in Korea (43.5%-54%) (14, 16, 17). These might be due to difference of the demographic factors of the enrolled patients, such as their gender and smoking status. More than half of the patients in the former studies were non-smokers and women, in contrast to the study population included in the present investigation (14, 16). Since lung adenocarcinomas with solid component comprised of a large proportion of the sub-histology in our study compared with other researches, activating EGFR mutations were less frequently seen (14). The characteristics of the regional population are also important factors affecting the incidence of activating EGFR mutations in lung adenocarcinomas even in the same country. It is necessary to analyze the characteristics of the patients in each study.
In our study, the incidence of EGFR mutation among male patients who smoked is relatively high (26.0%, 50 out of 192) compared to male patients in Spain (8.2%) or smokers in Italy (15.3%) (18, 19). In another Korean study, the incidence of EGFR mutations was 34.3% in men who smoked (14). In another report investigated in Asians, the incidence was 44.0% in male and 41.2% in ever-smokers (20). Although racial differences and the various inclusion criteria of each study might be the cause of differences in these results, the discordance of EGFR mutations in male smokers is not well-established. Considering that male smokers comprise the largest proportion of patients with NSCLC, a frequency of about one out of four patients is not a small number. Therefore, small biopsy specimens from patients with advanced lung cancer, even if the patient is a heavy smoker, should be used for EGFR mutation analysis in Asians, including Koreans.
According to recently published articles, the patient's age independently increased the likelihood of harboring an EGFR mutation (17, 21, 22). In contrast, the mean age of the EGFR-mutant group was younger than the EGFR wild-type group in our study. It seems that our study population consisted of older patients, included relatively more heavy smokers and male patients. The mean age of light smokers was younger than that of heavy smokers in this research (64.4 vs. 66.7, p=0.051).
There was no difference in the frequency of EGFR mutation according to the diagnostic method. This means that the location of the cancer did not influence the likelihood of having an EGFR mutation. However, the clinical stage was a related factor. It seemed to be more likely an EGFR mutation was present when NSCLC was diagnosed at an earlier stage. This contradicts the result of previous work in which patients presenting with miliary intrapulmonary carcinomatosis pattern at initial diagnosis had high rates of adenocarcinoma and EGFR mutation (23). These discrepant results may be due to the small sample size in both investigations.
The present study showed that the frequency of activating EGFR mutation was different between histological subtypes of lung adenocarcinoma. Recently, several studies have look for association between carcinoma histological subtypes and oncogenic driving mutations (14, 22, 24, 25). Consistently, EGFR-mutant tumors were more likely to be of the acinar subtype and less likely to be solid or mucinous. This study supported the notion that the presence of any amount of lepidic component correlated positively with the presence of activating EGFR mutation (24). While a significant correlation between micropapillary subtype and positive EGFR mutation was reported by several studies, no significant relation was seen in this study mainly because of the limited specimens with micropapillary components (10.4%, 14/135) (14, 24). Classification of the sub-histology of lung adenocarcinoma in small biopsy specimens is relevant because particular subtypes were associated with EGFR mutations and it provides prognostic information (24, 25).
Also known as acquired EGFR-TKI-resistant mutation, the T790M mutation in exon 20 was found in two patients who were naïve to EGFR-TKI. This was in good agreement with previous findings that direct DNA sequencing was able to detect T790M in approximately 2% of pre-TKI patients with NSCLC carrying sensitive EGFR mutation (26). It was reported that TKI treatment was less effective in those who have pre-treatment T790M mutations (27), but it was not evaluated in this study.
This retrospective study had several limitations. Firstly, the EGFR-TKI response rate is more related to EGFR-TKI-sensitive mutation than clinical predictors including being Asian, female, an ever-smoker, and having adenocarcinoma (28). Among patients harboring activating EGFR mutations, solid predominant subtype was related to poorer treatment response to EGFR-TKI than other subtypes (29). However, EGFR-TKI response rate according to adenocarcinoma subtype was not evaluated in this study because of the limited number of patients who received EGFR-TKI treatment. Further investigation is needed to determine the clinical outcome for each sub-histological group. Secondly, careful interpretation of the result is required because this result was based on data from a small sample size at a single center.
The incidence of EGFR mutations in lung adenocarcinoma is associated with gender and smoking history. This is already well-known, but most investigations were carried on resected specimens (6, 7, 11, 28). We confirmed the relationship between the clinicopathological features and EGFR mutations in small biopsy specimens of lung adenocarcinoma. Acinar, papillary and lepidic subtypes of lung adenocarcinoma, even if the histology was confirmed only from small biopsies, can also predict a high frequency of EGFR mutations. EGFR-TKI-sensitive mutation is a better predictor of response to EGFR-TKI treatment. When EGFR mutation analysis cannot be performed, the clinicopathological features of the lung adenocarcinoma can be a good predictor.
Since most advanced lung carcinomas are diagnosed by small biopsies and the treatment options for patients with advanced lung cancer are limited to conventional chemotherapy or EGFR-TKI with no hope for a curative operation, it is particularly important to verify that the clinicopathological features available from small biopsy specimens can be used to predict the EGFR mutation status. This necessity is even more acute when the specimen available is not enough for DNA extraction and a re-biopsy is impossible. The clinicopathological features from the small biopsy can acceptably predict the EGFR mutation status of a lung adenocarcinoma, but it cannot substitute for the mutation analysis.
In conclusion, the results of this study demonstrate that histological subtype and TTF1 expression of lung adenocarcinoma in small biopsies can be used to predict EGFR mutations. In Korean patients with lung adenocarcinoma, it is important to test EGFR mutational status in men regardless of smoking status. When the specimens are not adequate for DNA extraction and re-biopsy is not possible, the clinicopathological features of lung adenocarcinoma can be used to predict the EGFR mutation status.
Footnotes
-
This article is freely accessible online.
- Received February 2, 2014.
- Revision received April 13, 2014.
- Accepted April 15, 2014.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved