Introduction
Lung cancer is the leading cause of cancer-related deaths in males and second after breast cancer in females. Its incidence is high and has a global age-standardized rate (ASR) of 47.4/100,000 and 18.6/100,000 for males and females, respectively (1). In the Arab countries, the incidence of lung cancer ranks after prostate, colon or bladder cancer in males and after breast and cervical cancer in females (2). The ASRs are 13.44/100,000 and 2.91/100,000 for males and females, respectively, but the highest incidence is reported in Tunisia followed by Bahrain. However, the largest number of cases occurs in North African countries, Egypt (20.6%), Morocco (20.1%), Algeria (15.4%) and Tunisia (10%), reflective of their high population. As for Lebanon, lung cancer incidence ranks third after Tunisia and Bahrain in males and second after Bahrain in females (3). The reported ASRs for lung adenocarcinoma in Lebanon are 27.3/100,000 and 10.2/100,000 for males and females, respectively (4).
For several decades lung cancer has been treated as a single disease, yet with greater knowledge of its molecular biology lung cancer is now considered a heterogeneous disease. Most important was the discovery of the association of epidermal growth factor receptor (EGFR) gene mutations and anaplastic lymphoma kinase (ALK) gene fusion with lung adenocarcinoma histology. These genetic aberrations have been linked to specific ethnicities and gender in never-smokers and have shown a good response to tyrosine kinase inhibitors (5–9). Consequently, mutational analysis of EGFR and EML-ALK has become a cornerstone in the management of patients with lung adenocarcinoma (10). Moreover, downstream of the EGFR pathway is the Kirsten rat sarcoma viral oncogene, or v-Ki-ras2 homolog (KRAS), known to occur in a wide range of cancers, with the highest prevalence in pancreatic (~90%), colorectal (~60%) and primary lung adenocarcinomas (~60%), which can be mutated in lung cancer (11). When this mutation is detected, it is mutually exclusive of EGFR mutations but only rare and exceptional cases demonstrate both mutations (13), and a number of studies correlate its presence with resistance to anti-EGFR therapy and poor patient prognosis (11).
Lung cancer is notoriously correlated with smoking; therefore, the US and a number of European countries have implemented tobacco control programs and attained a decreased incidence worldwide (14,15). Unfortunately, in the Arab countries the incidence is increasing since smoking habits are increasing particularly among females and youth (3). Furthermore, in the hope of decreasing cancer-related mortality, almost all ethnic groups have been investigated for EGFR and KRAS gene mutation frequency in lung adenocarcinoma (16,17), but such a study has not been conducted in any Arab population including the Lebanese. Therefore, we aimed to evaluate the prevalence of KRAS and EGFR gene mutations in two Lebanese lung adenocarcinoma cohort groups and to correlate the findings with the clinical and pathologic features, including age, gender, smoking history and histological grade.
Materials and methods
The Institutional Review Boards of the American University of Beirut Medical Center and Hammoud Hospital University Medical Center approved this study. Both boards waived the need for written patient informed consents.
Patient selection
Patient cases (242) diagnosed clinically with primary lung adenocarcinoma (AC) and 150 with non-small cell carcinoma-not otherwise specified (NSCLC-NOS) were retrieved from the archives of the Pathology Departments at the American University of Beirut Medical Center and Hammoud Hospital University Medical Center in Lebanon from the year 2001 to 2010. Of the 150 NSCLC-NOS cases, 91 had an adequate tumor size (>1 mm) for further subtyping by immunohistochemical staining and 37 cases were diagnosed as ‘favoring primary lung AC’. Moreover, to ensure good quality and quantity of extracted DNA for mutational analysis, all 279 primary lung AC cases were then evaluated microscopically for tumor size. The cases with >200 tumor cells were included; thus, 106 primary lung AC cases were selected for EGFR and KRAS mutational analysis. The clinicopathologic information pertaining to patient age at diagnosis, gender, history of smoking, grade and TNM stage was retrieved for all cases, and none of the patients in this study received chemotherapy prior to specimen resection (lobectomy) or biopsy.
Immunohistochemistry
Three histological sections of 3-μm thickness were prepared from formalin-fixed paraffin-embedded (FFPE) tissue blocks of the NSCLC-NOS cases. Immunomarker detection was performed using antibodies against TTF-1 (thyroid transcription factor-1) (1:200 dilution; Novocastra; SPT24), napsin A (1:400 dilution; Novocastra; IP64) and p63 (Ready-to-use; Novocastra; 7JUL). Positive staining for TTF-1 and/or napsin A was used for identifying AC and p63 for squamous cell carcinoma. The immunomarkers were considered positive if at least 10% of the tumor cells stained. Cases with <10% staining and no focal areas of positive staining were interpreted as negative. Appropriate positive and negative controls were included. Immunostaining by TTF-1 of entrapped normal lung epithelium or napsin A of pulmonary macrophages was ignored. Immunohistochemistry was performed using the Leica Bond-Max autostainer (Leica Microsystems Inc., Buffalo Grove, IL, USA) with the manufacturer’s preset timed reagents.
Mutational analysis for KRAS and EGFR
FFPE tissue blocks of the included 106 lung adenocarcinoma cases were collected for KRAS and EGFR mutational analysis according to the algorithm (Fig. 1).
DNA extraction from FFPE
Eight 10-μm FFPE tissue ribbons were obtained from each block, and lysis was allowed overnight at 65°C with 1.5 μl proteinase K (20 mg/ml) and 400 μl cell lysis solution (Qiagen; 158906). DNA was extracted following a standard protocol and quantified using the NanoDrop ND-1000 spectrophotometer. The resultant DNA was stored at 4°C until utilized for reverse hybridization and multiplex real-time polymerase chain reaction (PCR) procedures.
Reverse hybridization for detection of KRAS mutations
Detection of KRAS mutations in exon 2, codon 12 and 13 and exon 3 codon 61 was performed by reverse hybridization using the KRAS 12/13/61 StripAssay® (ViennaLab Diagnostics GmbH, Vienna, Austria). This technique involves three steps: PCR amplification using biotinylated oligonucleotide primers, hybridization of the amplified products to a strip containing allele-specific oligonucleotide probes immobilized as an array of parallel lines and placed in a shaking water bath at 45°C and identification of bound biotinylated sequences using streptavidin-alkaline phosohatase and color substrates performed at room temperature. Bands were then analyzed visually. Cases that did not exhibit these mutations were then submitted for EGFR exons 18–21 mutational analysis, knowing that KRAS and EGFR mutations are predominantly mutually exclusive.
Multiplex real-time PCR for detection of EGFR mutations
EGFR mutational analysis was performed using the EGFR RGQ PCR kits (Qiagen, Valencia, CA, USA), that enables detection of 29 somatic mutations in the EGFR gene including: 19 deletions in exon 19, T790M-exon 20, L858R-exon 21, L861Q-exon 21, G719X-exon 18, S768I-exon 20 and 3 insertions in exon 20. The kit utilizes two technologies, amplification refractory mutation system (ARMS) technology and Scorpions dual primer probes available for detection of EGFR mutations using real-time PCR on the Rotor-Gene Q 5-plex HRM instrument (Qiagen). This analysis is performed in two steps: the first step is to perform a control assay in order to assess the total DNA in the sample and the second step is to complete the assay for the presence or absence of mutated DNA. This procedure was completed as per the manufacturer’s instructions manual. We considered this technology because of its high sensitivity particularly for samples with low levels of tumor DNA, such as in small lung biopsy samples.
Statistical analysis
Age, gender, smoking history, grade and TNM stage were summarized using means and standard deviation for age and frequency distribution for the other 4 variables. Those variables were compared between patients with KRAS mutation and those without KRAS mutation using the independent t-test for age (or Wilcoxon rank sum test for small sample size) and the Chi-squared test or Fisher’s exact test (when cell counts were small) for the other 4 variables. Similar comparisons were made between patients with an EGFR mutation and those without mutations.
Results
The overall clinical and pathological data of the 106 primary lung adenocarcinoma cases demonstrated a mean age of 62 years and a male to female ratio of 2:1. The majority of patients were smokers (55.7%), although we could not determine the smoking history data for 27.7%. A total of 56.2% of the tumors were poorly differentiated when applying the three-tier grading system. TNM staging data were available for 45% of the cases, which limited our statistical analysis, but more tumors had a size larger than 3 cm, while approximately half of the cases had lymph node and/or distant metastasis (Table I).
The prevalence of KRAS mutations was detected in 37.7% of the primary lung AC cases. The majority (85%) had a G>T substitutions in codon 12 of exon 2 and 6 cases had this substitution in codon 13 (Table II). An A>G substitution in codon 61 of exon 3 was detected in one case. All of the KRAS mutations were single except for 5 cases that exhibited double mutations. The KRAS mutations were predominant in males (67.5 vs. 32.5% in females) and in smokers (57.5 vs. 10.0% in non-smokers); however, no statistical significance could be concluded (p=0.942 and p=0.286, respectively). The majority of the KRAS-mutated tumors were poorly differentiated (62.5%) and had a tumor size larger than 3 cm, but the frequency of lymph node or distant metastasis was not significantly higher when compared to the lung adenocarcinoma cases with no KRAS mutation (Table I).
EGFR mutations were detected in 9 (8.5%) of the lung adenocarcinoma patients with no KRAS mutations. The most common mutations (8 cases, 88.9%) were deletion in exon 19 while one case had a substitution L858R in exon 21 (Table III). Predominance of EGFR mutations was significant in females (p=0.005), non-smokers (p=0.003) and well differentiated tumors (p<0.001). TNM data were not available for all the 9 cases; therefore, we could not draw a significant correlation (Table I).
In summary, we identified mutations in 46.2% of the lung adenocarcinoma cases distributed into 37.7% with KRAS mutations and 8.5% with EGFR mutations (Table IV).
Discussion
Molecular genetic analysis of lung adenocarcinoma has become an integral part of lung cancer diagnosis and present treatment (18,19). The most commonly detected driver mutations are those involving KRAS, EGFR and EML4-ALK genes. Moreover, EGFR tyrosine kinase inhibitors (TKIs), gefitinib, erlotinib and afatinib, are considered as first-line targeted therapy for advanced (locally or metastatic) lung adenocarcinoma. However, the 3prevalence of these aberrations varies with gender, ethnicity and smoking history (20). Consequently, lung adenocarcinomas have been extensively profiled for genetic aberrations in most ethnic groups. In the Arab population, only one study evaluated EGFR genetic aberrations in 34 Saudi lung cancer patients (21). In the present study, 106 cases of lung adenocarcinomas from Lebanon were analyzed for mutations in KRAS and EGFR.
EGFR mutations were detected in 8.5% of the lung adenocarcinomas, a frequency higher than that in the Saudi population, comparable to that of some European countries, but much lower than that of Asian countries. The highest reported frequency of EGFR mutations is in Asian, non-smoker females (Taiwan, 57.3%) (22). However, the Europeans reported much lower frequencies and Szumera-Ciećkiewicz et al reviewed the results from 19 European studies and showed a range of 2.6% (in Italy) to 39% (in Germany) (23). In the US, a study of 3026 cases by Dogan et al demonstrated a frequency of 20% (24) and Reinersman et al reported a frequency of 19% in African-Americans (25). In Latin America, the average frequency is 33.2% with the lowest in Argentina (19.3%) and the highest in Peru (67%). This reported frequency is correlated with the high rate of Asian migration (26). To the best of our knowledge, in the Middle East and the Levant countries, only one study from Saudi Arabia has been published. This study applied direct DNA sequencing to 34 lung cancer cases and detected the EGFR mutation in one lung adenocarcinoma (3%) (21).
The analysis of EGFR mutation subtypes demonstrated a predominance of deletions in exon 19 and only one of the 9 cases had an L858R substitution in exon 21. These findings are comparable to the reported frequency of EGFR mutation subtypes: in-frame deletions in exon 19 (44% of all mutations), missense point mutations (L858R substitution) in exon 21 (41% of all mutations), in-frame duplications/insertions in exon 20 (5% of all mutations) and G719X (X indicates A, C or S) substitution in exon 18 (4% of all mutations) (27). It is important to subtype these mutations since they respond differently to EGFR TKIs (28). For example, the T790M mutation in exon 20 correlates with drug resistance and relapse (29). In addition, we demonstrated a significant correlation between EGFR mutations and females, non-smokers and well differentiation of the tumor. These findings are in concordance with the reported prevalence of these mutations in females more than in males and in never-smokers than in smokers and to their correlation with a better prognosis (30).
KRAS mutations were detected in 37.7% of the lung adenocarcinomas. Similar to EGFR mutations, the prevalence of KRAS mutations in lung adenocarcinoma varies with ethnic groups as well, but is generally mutually exclusive with EGFR mutations (13,31). An overview of KRAS and EGFR mutation frequency from different countries presented in Table V demonstrates an inverse EGFR mutation frequency as compared to that of KRAS. The lowest values were reported in Asians (Taiwan: KRAS, 5.03%; EGFR, 57.3%), while the highest KRAS prevalence was reported in Europeans (The Netherlands: KRAS, 36.9%; EGFR, 10.6%).
In addition, we found that KRAS mutations were predominantly in exons 2 and 3 with 85% involving codon 12, thus concordant with that reported by Bos, where 97% of KRAS mutation subtypes are substitutions of G→T and G→C in codon 12 of exon 2 in addition to the missense point mutations in codons 61 that prevent GTP cleaving, thus transforming the gene into an oncogene (32). However, we could not demonstrate a correlation between smoking and the type of substitution and the subtype of the mutation. Such a correlation was described by Riely et al where they demonstrated that the G→T and G→C transversions are present in both smokers and non-smokers, while a G→A transversion is more common in non-smokers (12). Moreover, we detected more KRAS mutations in males, smokers and poorly differentiated tumors, but this observation did not demonstrate a statistically significant correlation. Several other studies have shown that KRAS mutations are higher in males and smokers as compared to females and never-smokers (33), predict a worse prognosis (34,35), and show a low response to adjuvant chemotherapy and EGFR-TKI treatment in metastatic lung adenocarcinoma (7). Unfortunately, because it leads to loss of enzymatic activity, effective targeted therapy is more difficult to achieve as compared to a gain of functional mutation. Currently, clinical trials targeting downstream effector molecules are being conducted, but no effective targeted therapy against KRAS mutations has been achieved thus far.
Two mutational analysis techniques were used in this study. The Scorpion amplified refractory mutation system (ARMS) multiplex real-time PCR was used for the detection of EGFR mutations. Although this technique does not detect mutations comprehensively, it covers all the common 29 EGFR mutations in exons 18–21. It is highly sensitive (1%) as compared to direct Sanger sequencing (25%) of mutant DNA (36). For KRAS mutations we used reverse hybridization (StripAssay, ViennaLab), which is considered most analytically sensitive when compared to direct sequencing, pyrosequencing, high resolution melting analysis and the TheraScreen DxS kit (37). This technique can detect 13 mutations in the KRAS codon 12, 13 and 61. The high sensitivity of both techniques was necessary to overcome the tissue limitation of small lung biopsy material that constituted the majority of our cases. However, some biopsy material was too scant to be included in this study thus constituting a limitation for this study. In addition, the unavailability of adequate smoking history data for all cases, and being retrospective in nature are additional limitations.
In summary, this is the first study in an Arab population to report the prevalence of both EGFR and KRAS gene mutations in lung adenocarcinoma using sensitive mutational analysis techniques. We showed that the frequency of EGFR in lung adenocarcinoma was 8.5% and that it was more common in women, non-smokers and well-differentiated tumors. In addition, we found that KRAS was highly prevalent in the Lebanese population, but we could not significantly correlate it with smoking; however, it remains a possible major cause. Therefore, we conclude that EGFR reflex testing should be implemented as per the recommendation of the College of American Pathologists, International Association for the Study of Lung Cancer and Association for Molecular Pathology (10). While KRAS testing is useful for understanding the molecular biology of adenocarcinoma, the clinical utilization of this information must await the identification of effective targeted therapy.