Background

Lung cancer is the leading cause of cancer-related mortality worldwide [1]. Non-small cell lung cancer (NSCLC) is the most common histological subtype of lung cancer, accounting for approximately 80–85% of lung cancer cases [2, 3]. Molecular testing for epidermal growth factor receptor gene (EGFR) mutations and ALK receptor tyrosine kinase (ALK) translocations has become the evidence-based standard of care for the management of advanced NSCLC. In the past, pivotal clinical trials have demonstrated clinical benefit from targeting EGFR mutations and ALK translocations, and currently a number of effective EGFR and ALK inhibitors are available for targeted therapy of NSCLC harboring the relevant aberrations [4]. More recently, new molecular profiling technologies have permitted the identification of other potential oncogenic drivers including mutations in the KRAS proto-oncogene (KRAS), B-Raf proto-oncogene (BRAF), erb-b2 receptor tyrosine kinase 2 gene (ERBB2), MET proto-oncogene (MET) and phosphatidylinositol-3 kinase catalytic subunit alpha gene (PIK3CA) as well as ROS proto-oncogene 1 (ROS1) and ret proto-oncogene (RET) rearrangements [4]. While a number of studies have already evaluated the frequencies of these genetic alterations in NSCLC patients from different countries, information on the prevalence of oncogenic driver mutations in the Swiss population are scarce and limited to population based epidemiological data derived from cancer registries and molecular test results based exclusively on Sanger sequencing rather than next generation sequencing (NGS) [5, 6].

In Switzerland lung cancer is the most common cause of cancer-related death among men (approximately 2000 deaths per year) and the second most common cause of cancer-related death among women (approximately 1100 deaths per year) [7]. Adenocarcinoma is the predominant histological subtype with distinct molecular features, and incidence rates of lung adenocarcinoma are increasing among both sexes [8, 9]. The aim of the study was to analyze the frequencies of ALK, RET and ROS1 gene rearrangements and EGFR, KRAS, BRAF, ERBB2, MET and PIK3CA mutations in a representative cohort of Swiss patients with lung adenocarcinoma using NGS as testing method in the majority of cases and to correlate the molecular findings with clinicopathological patient characteristics.

Methods

Patients

A total of 475 consecutive patients who underwent molecular testing of newly diagnosed lung adenocarcinoma at the Institute of Pathology and Molecular Pathology, University Hospital Zurich (Zurich, Switzerland), between January 2014 and January 2018, were included in the study, independent of tumor stage. Molecular analyses were performed at the University Hospital Zurich according to National Comprehensive Cancer Network (NCCN) and Swiss Society of Pathology (SSPath) guidelines. Inclusion criteria were histologically and/or cytologically confirmed lung adenocarcinoma, chemotherapy, targeted therapy and radiotherapy naïve, and tissue blocks/cell blocks with adequate tumor cellularity. Exclusion criteria were non-adenocarcinoma histology, previous chemotherapy, targeted therapy or radiotherapy, and insufficient tumor material. Of the initial study population, 469 patients had adequate tumor material for molecular testing, while 6 patients had insufficient tumor samples and were not further evaluated. The results of molecular analysis were recorded for each patient and correlated with demographic and tumor related data such as gender, age, smoking status, clinical stage, and TNM stage (as defined by the Union for International Cancer Control (UICC) TNM classification of malignant tumors, 8th edition [10]). Smoking status was defined as never smokers (< 100 lifetime cigarettes), ex-smokers (≥100 lifetime cigarettes and currently not smoking) and current smokers (≥100 lifetime cigarettes and currently smoking). The cutoff date for data collection was 15 May 2018. The study was approved by the Cantonal Ethics Committee of Zurich (StV-No. 2009/14–0029).

Molecular analysis

Nucleic acids (DNA and RNA) were isolated from formalin-fixed paraffin-embedded (FFPE) tissue blocks or FFPE cell blocks using the Maxwell 16 FFPE Tissue LEV DNA/RNA Purification Kit (Promega, Fitchburg, WI, USA). The obtained nucleic acids were quantified with NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA) and Qubit 2.0 (Thermo Fisher Scientific/Life Technologies, Eugene, OR, USA) using the dsDNA/RNA HS Assay Kit (Thermo Fisher Scientific/Life Technologies, Zug, Switzerland). Mutation analysis was performed using Sanger sequencing (n = 158) or NGS (n = 311). For DNA- and RNA-based NGS, customer panels including the Ion AmpliSeq Colon and Lung Cancer panel 2 (CLP2), Ion AmpliSeq Fusion Lung Cancer Research panel (LFP), and Oncomine DNA panel for Solid Tumors and Fusion Transcripts (Thermo Fisher Scientific/Life Technologies, Carlsbad, California, USA) were applied, as previously described [11, 12]. Briefly, we used the Ion Library Quantitation kit (Thermo Fisher Scientific) for quantification of DNA and RNA libraries, the Ion One Touch 200 Template Kit v2 DL (lately replaced by the Ion Hi-Q Chef Kit and the Ion Chef System) (Thermo Fisher Scientific) for emulsion polymerase chain reaction (PCR) and template preparation, and the Ion Personal Genome Machine 200 Kit v2 (lately replaced by the Ion Personal Genome Machine Hi-Q Sequencing Kit) (Thermo Fisher Scientific) as the sequencing platform. For Sanger sequencing, we used the Illustra GFX PCR DNA and Gel Band Purification Kit (GE Healthcare Life Sciences, Buckinghamshire, UK) for purification of amplified DNA fragments, the Genetic Analyzer 3130 × 1 (Applied Biosystems, Foster City, CA, USA) for sequencing and the Sequencher 5.1 (Gene Code Corporation, Ann Arbor, MI, USA) for data analysis. ALK and ROS1 immunohistochemistry (IHC)/immunocytochemistry (ICC) was performed on the automated immunostainer DiscoveryUltra (Roche Ventana) using a mouse anti-human ALK monoclonal antibody (clone 5A4, Leica Biosystems) and a rabbit anti-human ROS1 monoclonal antibody (clone D4D6, Cell Signaling Technology). ALK or ROS1 IHC/ICC positive cases were confirmed by fluorescence in situ hybridization (FISH) using the Vysis LSI ALK Dual Color Break Apart Rearrangement Probe (Abbott Molecular, Baar, Switzerland) and the ZytoLight SPEC ROS1 Dual Color Break Apart Probe (Zytovision GmbH, Bremerhaven, Germany). FISH testing for RET rearrangement was performed using the ZytoLight SPEC RET Dual Break Apart Probe (Zytovision GmbH, Bremerhaven, Germany). For each case, a board certified pathologist analyzed 50–100 tumor nuclei. A sample was considered positive, if split signals were detected in ≥15% of tumor nuclei according to the manufacturer’s evaluation guidelines (Abbott Molecular, Des Plaines, IL, USA).

Statistical analysis

Descriptive statistics were employed to describe the patient characteristics of the study cohort. The results are presented as frequencies and percentages for categorical variables and as mean ± standard deviation, median and range for continuous variables. Associations between mutation status and clinicopathological characteristics were tested using univariate and multivariate analyses. Univariate analysis was performed by chi-square test or Fisher exact test for categorical variables and by t test or nonparametric Mann-Whitney test for continuous variables. Multivariate analysis was performed by logistic regression. P-values < 0.05 were considered statistically significant. All statistical analyses were performed using SPSS Statistics software (version 24.0, IBM, Ehningen, Germany).

Results

The diagnosis of lung adenocarcinoma was based on histology (with or without cytology) in 91.7% (430/469) and on cytology alone in 8.3% (39/469) of the patients. Samples submitted for molecular testing were obtained from primary tumors, lymph node metastases or distant metastases in 79.7% (374/469), 10.7% (50/469) and 9.6% (45/469) of the patients, respectively. There were 191 (40.7%) resection specimens, 224 (47.8%) biopsy specimens, 48 (10.2%) fine needle aspiration/bronchial brushing/bronchoalveolar lavage specimens and 6 (1.3%) cell blocks from pleural effusions. Table 1 summarizes the demographic and clinicopathological patient characteristics. The study population consisted of 235 men and 234 women (mean age at diagnosis, 64.1 ± 11.4 years; range, 27–94 years). The majority of patients were ever smokers (current smokers and ex-smokers) (354/469, 75.5%) and had clinical stage IV lung adenocarcinoma (299/469, 63.8%) at diagnosis. Females were more likely to be never smokers than males (70/234, 29.9% vs 45/235, 19.1%, p = 0.007, beta 0.589, OR 1.802, CI 95% 1.174–2.767). Overall 127 patients received targeted treatment. Stage IV patients (both at diagnosis and during follow-up) with EGFR mutation and ALK rearrangement received targeted treatment in 75.4% and 61.9%, respectively. The majority (91.7%) of stage IV patients with EGFR mutation who did not receive targeted therapy were treated with chemotherapy and/or radiotherapy. Likewise, all stage IV patients with ALK translocation who were not treated with targeted therapy received chemotherapy and/or radiotherapy. Median patient follow-up was 17 months (range, 1–52 months). 268/469 (57.1%) patients were alive at the time of last follow-up, including 165/469 (35.2%) patients with stable disease and 103/469 (22.0%) patients with progressive disease. 147/469 (31.3%) patients died of disease during follow-up, and 54/469 (11.5%) patients were lost to follow-up. Median overall survival for the entire study cohort was 38 months.

Table 1 Patient characteristics
Full size table

EGFR mutation analysis

A total of 95 EGFR mutations were detected in 90/469 (19.2%) patients. The most common EGFR mutations were exon 19 deletions (49/90, 54.4%, most frequent subtype: E746_A750del, 33/90, 36.7%) and exon 21 L858R missense mutations (35/90, 38.9%) (Table 2). Doublet EGFR mutations were found in 5 (5/90, 5.6%) tumors, including 2 tumors with L858R and non-L858R missense mutations, 2 tumors with two non-L858R missense mutations and 1 tumor with non-L858R missense mutation and T854A primary resistant mutation (Additional file 1: Table S1). The analysis of the distribution of EGFR mutations among men and women showed a predominance of EGFR mutations in the female group: 55/234 (23.5%) women had a total of 59 EGFR mutations (most frequent: exon 19 deletion, 31/55, 56.4%). By contrast, 35/235 (14.9%) males had a total of 36 EGFR mutations, the most common of which – as in the female group – were exon 19 deletions (18/35, 51.4%). The highest prevalence of EGFR mutations was observed in never smokers (69/115, 60.0%) and was considerably lower in ex-smokers (10/160, 6.3%) and current smokers (11/194, 5.7%). The association between EGFR mutational status and either gender or smoking status was statistically significant on univariate (p = 0.019 and p < 0.001, respectively) and multivariate analyses (p = 0.033 and p < 0.001, respectively) (Tables 3 and 4). No statistically significant differences were found between EGFR mutated and EGFR wildtype tumors with respect to clinical stage (except for stage III, p = 0.040), T stage (except for T2a, p = 0.001), N stage, M stage (except for M1c, p = 0.011), tumor location, mean tumor size and mean patient age (Table 3).

Table 2 EGFR mutations in 90 lung cancers
Full size table
Table 3 Associations between clinicopathological features and EGFR mutational status
Full size table
Table 4 Logistic regression analysis of EGFR mutational status
Full size table

KRAS mutation analysis

Of 443 patients with informative KRAS mutation analysis, 159 (35.9%) harbored KRAS mutations. KRAS mutations were most frequently located in exon 2 (154/159, 96.9%), and the most common mutations were G12C (72/159, 45.3%) and G12 V (26/159, 16.4%) (Table 5). Of 443 patients with informative KRAS and EGFR mutation analyses, 2 patients (0.5%) had coexistent KRAS and EGFR mutations (one with G13S and E746_A750del and one with G12 V and E709A). KRAS mutations tended to be more frequent in females (86/217, 39.6%) than in males (73/226, 32.3%) and were more commonly found in ever smokers (152/353, 43.1%) than in never smokers (7/90, 7.8%). The association between KRAS mutation and smoking status was statistically significant on both univariate and multivariate analyses after stratification by gender (p < 0.001 and p < 0.001, beta −2.285, OR 0.102, CI 95% 0.045–0.228), while gender was significantly associated with KRAS mutation only on multivariate analysis after adjusting for smoking (p = 0.016, beta 0.507, OR 1.660, CI 95% 1.099–2.507) (Tables 6 and 7). Among the patients with informative KRAS mutation analysis, males were significantly more likely to be ever smokers (current smokers or ex-smokers) than females (190/226, 84.1% vs 163/217, 75.1%, p = 0.020, beta 0.559, OR 1.748, CI 95% 1.092–2.800). No statistically significant differences were found between KRAS mutated and KRAS wildtype tumors with respect to clinical stage, T stage (except for T2, p = 0.043), N stage, M stage, tumor location (except for the left lower lobe, p = 0.006), mean tumor size and mean patient age (Table 6).

Table 5 KRAS mutations in 159 lung cancers
Full size table
Table 6 Associations between clinicopathological features and KRAS mutational status
Full size table
Table 7 Logistic regression analysis of KRAS mutational status
Full size table

ALK rearrangement analysis

ALK rearrangement was detected by FISH (n = 20) or NGS (n = 8) in 28/376 (7.4%) tumors, including one with coexistent KRAS mutation (G12 V). Of the 8 cases with ALK rearrangement diagnosed by NGS, EML4 exon 13-ALK exon 20 fusion gene variant was found in 4 (50.0%) cases, EML4 exon 6-ALK exon 20 fusion gene variant was detected in 3 (37.5%) cases, and EML4 exon 18-ALK exon 20 fusion gene variant was detected in 1 (12.5%) case. There was no significant difference in the frequency of ALK rearrangement between males and females (15/198, 7.6% vs 13/178, 7.3%, univariate analysis, p = 0.920, multivariate logistic regression, p = 0.669) (Tables 8 and 9). By contrast, ALK rearrangement was significantly more common in never smokers than in ever smokers (12/84, 14.3% vs 16/292, 5.5%, p = 0.007) (Table 8). The association between ALK rearrangement and smoking status remained statistically significant on multivariate analysis after adjusting for gender (p = 0.008, beta 1.081, OR 2.948, CI 95% 1.323–6.567) (Table 9). Among the patients tested for ALK rearrangement, females were more likely to be never smokers than males (49/178, 27.5% vs 35/198, 17.7%, p = 0.023, beta 0.570, OR 1.769, CI 95% 1.082–2.892). No statistically significant differences were found between ALK-rearranged and ALK wildtype tumors with respect to clinical stage (except for stage III, p = 0.038), T stage (except for T1b, p = 0.025), N stage (except for N2, p = 0.003), M stage (except for M1b, p = 0.013), tumor location (except for the right upper lobe, p = 0.001, and the middle lobe, p = 0.019), mean tumor size and mean patient age (Table 8).

Table 8 Associations between clinicopathological features and ALK rearrangement
Full size table
Table 9 Logistic regression analysis of ALK rearrangement
Full size table

KRAS mutation analysis was not performed in 26 patients with proven EGFR mutation, and ALK rearrangement testing was not performed in 93 patients including 62 patients with KRAS mutation and 31 patients with EGFR mutation. Because genetic alterations in EGFR, KRAS and ALK are generally mutually exclusive, it can be concluded that 90/469 (19.2%) patients had EGFR mutations, 159/469 (33.9%) had KRAS mutations, and 28/469 (6.0%) had ALK gene rearrangement (Table 10). 195/469 (41.6%) patients had triple-negative (EGFR-negative/KRAS-negative/ALK-negative) lung adenocarcinomas.

Table 10 Frequency of oncogenic driver mutation in our study cohort
Full size table

EGFR, KRAS and ALK comparative analyses

Comparative analyses of EGFR and KRAS mutated tumors, EGFR mutated and ALK rearranged tumors and KRAS mutated and ALK rearranged tumors are summarized in Additional file 1: Tables S2–S4. Of note, EGFR mutated tumors were more likely to have multiple extrathoracic metastases (M1c) compared with KRAS mutated tumors (37/90, 41.1% vs 44/159, 27.7%, p = 0.030) (Additional file 1: Table S2). EGFR mutated tumors were also more likely to be clinical stage IV and have single or multiple extrathoracic metastases (M1b or M1c) compared with ALK rearranged tumors (65/90, 72.2% vs 14/28, 50.0%, p = 0.029 and 49/90, 54.4% vs 7/28, 25.0%, p = 0.006) (Additional file 1: Table S3). ALK rearranged tumors were more frequently associated with clinical stage III than EGFR mutated tumors (10/28, 35.7% vs 12/90, 13.3%, p = 0.008) and more commonly showed ipsilateral mediastinal or subcarinal lymph node metastasis (N2) compared with EGFR and KRAS mutated lung adenocarcinomas (15/28, 53.6% vs 20/90, 22.2%, p = 0.002 and 15/28, 53.6% vs 48/159, 30.2%, p = 0.016).

Other mutations and rearrangements

RET fusions were detected by FISH in 4 (1.9%) of 208 tested patients, including 1 patient with coexistent PIK3CA mutation (E545K). ROS1 fusions were detected by FISH (n = 5) or NGS (n = 3) in 8 (3.2%) of 248 tested patients. No statistically significant differences were found between RET/ROS1 rearranged and non-rearranged tumors with respect to gender, smoking status, clinical stage, TNM stage, tumor location, mean tumor size and mean patient age. 12/309 (3.9%) tumors harbored BRAF mutations. The majority of BRAF mutations were located in exon 15 (10/12, 83.3%); the most common BRAF mutation was V600E (9/12, 75.0%) (Table 11). No statistically significant differences were found between BRAF mutated and BRAF wildtype tumors with respect to the clinicopathological parameters evaluated. ERBB2 mutations were detected in 9/286 (3.1%) tumors (Table 12), including 7 insertion/duplication mutations in exon 20, 1 nonsense mutation in exon 13 and 1 missense mutation in exon 8 of the ERBB2 gene; the most frequent ERBB2 mutation was p.A775_G776insYVMA (alternative nomenclature p.Y772_A775dup, c.2313_2324dup) (5/9, 55.6%). ERBB2 mutations were more common in never smokers than in ex−/current smokers (5/64, 7.8% vs 4/222, 1.8%, chi-square test, p = 0.029, multivariate logistic regression, p = 0.020, beta 1.621, OR 5.059, CI 95% 1.296–19.747), while no significant differences were found between ERBB2 mutated and ERBB2 wildtype tumors with respect to the other clinicopathological parameters analyzed. Nine MET exon 14 skipping mutations were detected in 234 (3.8%) tumors, including one with coexistent BRAF (V600E) and PIK3CA (E542K) mutation and one with coexistent KRAS (G13C) mutation. PIK3CA mutations were detected in 7/163 (4.3%) tumors (3 x E542K, 2 x E545K, 1 x R38H, 1 x H1047R), including one with coexistent BRAF (V600E) and MET exon 14 skipping mutation, two with coexistent EGFR mutations (L858R and L747_P753delinsS, respectively), one with coexistent KRAS (G12A) mutation and one with coexistent RET rearrangement. No statistically significant differences were found between MET/PIK3CA mutated and MET/PIK3CA wildtype tumors with respect to gender, smoking status, clinical stage, TNM stage, mean tumor size and mean patient age. While MET mutated tumors were more likely to be located in the right upper lobe than MET wildtype tumors (6/9, 66.7% vs 50/225, 22.2%, p = 0.007), PIK3CA mutated tumors were less likely peripheral in location and involved more frequently both the central and peripheral portions of the lung compared with PIK3CA wildtype tumors (2/7, 28.6% vs 107/156, 68.6%, p = 0.041 and 3/7, 42.9% vs 10/156, 6.4%, p = 0.012). Among the 469 study patients, 154 (32.8%) had lung adenocarcinomas that were negative for all oncogenic driver mutations evaluated in the current study.

Table 11 BRAF mutations in 12 lung cancers
Full size table
Table 12 ERBB2 mutations in 9 lung cancers
Full size table

Discussion

This study presents for the first time data on the EGFR, KRAS, ALK, ROS1, RET, BRAF, ERBB2, MET and PIK3CA mutation frequencies in a representative Swiss cohort of patients with stage I-IV lung adenocarcinoma using NGS as testing method in the majority of patients. Molecular testing was performed in all patients at the time of initial diagnosis during a 4-year period at a primary referral center for lung diseases in Northeastern Switzerland. We also comprehensively studied types of mutations and associations of mutational status with demographic and clinicopathological patient characteristics.

The reported EGFR mutation rate in patients with lung adenocarcinoma varies widely between different populations worldwide, ranging from 10 to 20% in European and North American cohorts [5, 6, 13,14,15,16,17,18,19,20,21,22,23] to more than 50% in Asian populations [24, 25]. The wide range of reported EGFR mutation rates among European cohorts might be explained by differences between the published studies with respect to patient selection criteria and methods used for molecular analysis. In a French study by Vallee et al. [19], one of the largest single center studies in Europe, EGFR mutations were detected in 13.5% of patients with NSCLC and in 14.7% of patients with lung adenocarcinomas. The authors used allele-specific PCR for evaluation of L858R point mutation and DNA fragment analysis to detect exon 19 deletions. Because other EGFR mutations were not evaluated, the true prevalence of EGFR mutations in this study remains unknown. The INSIGHT study, a large multicenter study comprising 1785 NSCLC patients (including 1393 patients with lung adenocarcinoma), showed an EGFR mutation frequency of 13.8% in NSCLC patients and of 15.4% in patients with lung adenocarcinoma [14]. The study analyzed tumor samples from 14 cancer centers in six Central European countries, each with different patient inclusion criteria and testing methods, which makes comparison with other studies more difficult. In addition, mutation testing was not performed at a fixed time point, which could induce bias as mutations may arise during the disease course [23]. Our study results show a prevalence of EGFR mutations that is similar to that reported by Moiseyenko et al. [20] (19.8%) in a Russian cohort and by Hlinkova et al. [22] (20%) in a Slovakian cohort, but lower than the EGFR mutation rates reported in two previous studies from Switzerland [5, 6]. Ess et al. [5] retrospectively analyzed population based data on the frequency of molecular testing, factors affecting testing and the prevalence of EGFR mutations and ALK rearrangements in patients with stage IV or relapsed non-squamous NSCLC (including adenocarcinoma, large cell carcinoma and NOS histology) from 2008 to 2014. Using direct sequencing (EGFR exons 18–21) for EGFR mutation analysis and FISH with a break-apart probe for ALK rearrangement testing, EGFR mutations (exclusively exon 19 deletions and exon 21 L858R mutations!) were detected in 11% of patients with advanced non-squamous NSCLC and in 13% of patients with lung adenocarcinoma, while 12% of patients with non-squamous NSCLC and 10% of patients with lung adenocarcinoma harbored ALK rearrangements. Other oncogenic driver mutations or associations between EGFR mutation/ALK rearrangement status and clinicopathological characteristics of patients with lung adenocarcinoma were not evaluated. More recently, Schwegler et al. [6] prospectively analyzed population based epidemiological data on overall survival of patients with mutated stage IV lung adenocarcinoma, mostly residents in rural areas of Central Switzerland, from 2010 to 2014. EGFR mutations were detected with Sanger sequencing in 14% of the patients, while KRAS, ERBB2, BRAF and MET mutations and ALK and RET translocations were found in 20%, 2%, 1%, 0.5%, 6% and 0.5%, respectively [6]. In contrast to our study, the types of mutations were not analyzed, and mutational status was not correlated with demographic or clinicopathological features. Possible reasons for the reported lower EGFR mutation rates compared with that of our study may be different modes of patient selection (selection from the molecular database of University Hospital Zurich vs selection from cancer registries), different patient selection criteria (patients with stage I-IV lung adenocarcinoma vs patients with stage IV or relapsed non-squamous NSCLC [5] and patients with stage IV lung adenocarcinoma [6]) and different methods used for mutational analysis (NGS and Sanger sequencing vs Sanger sequencing alone). In contrast to the studies by Ess et al. [5] and Schwegler et al. [6], the majority of patients in our study underwent molecular testing with NGS, which has been shown to demonstrate high analytic sensitivity, accurate detection of complex indel mutations, and broad reportable ranges with simultaneous detection of doublet EGFR mutations and concomitant KRAS and BRAF mutations in the clinical diagnostic setting [2, 26]. In addition, in the study by Schwegler et al. [6] patients with stage I-III lung adenocarcinoma were excluded from the analysis, and 20% of stage IV lung cancer patients were not tested for oncogenic driver mutations, while in the study by Ess et al. [5] 38% of patients did not receive molecular analysis. Although we did not assess mutation testing rates at our institution, it can be assumed that the molecular testing rates in the period from 2014 to 2018 were higher than those of previous years and that patients treated at an institution active in clinical research are more regularly tested for predictive biomarkers than patients treated at an institution not participating in clinical research [5]. In accordance with published literature [13,14,15,16, 23], we found a significant association of EGFR mutation status with female gender and never smoking status. When we restricted the analysis to female never smokers, we achieved a high EGFR mutation rate of 65.7% (46/70), a finding consistent with previous reports [13, 24, 25].

KRAS mutation is one of the most frequent mutations in NSCLC, at least in Caucasian populations, with reported frequencies reaching up to 30% of lung adenocarcinomas [13, 23, 27, 28], while its prevalence in Asian populations is approximately 10% [29,30,31]. KRAS mutations are predominantly found in smokers [32], but they may occur in up to 15% of non-smokers [27]. To date, no effective anti-KRAS agent has been released, although a number of preclinical studies and clinical trials are currently underway, exploring novel therapeutic approaches to target KRAS mutated NSCLC [33,34,35,36]. The KRAS mutation rate in our study was slightly higher than was previously reported for Caucasian populations (which might be related to different smoking habits in this Swiss cohort), but was almost identical to the KRAS mutation rate reported by Brcic et al. [13] in a Croatian cohort. The presence of KRAS mutation in our study was significantly associated with a history of smoking on both univariate and multivariate analyses, while the association of KRAS mutation with gender was statistically significant only on multivariate analysis after adjusting for smoking. This finding adds to a mixed body of literature. Some studies have shown increased incidence of KRAS mutations among females [23], while others found equal frequencies in both men and women [13, 37, 38].

ALK rearrangements are detected in 3–7% of NSCLC [39,40,41,42,43,44]. They predominantly occur in non-smokers, lung adenocarcinomas and non-Asian vs Asian populations, while men and women seem to be equally affected [3]. The frequency of ALK rearrangement in our study is consistent with previous reports, as is the association with smoking status (higher frequency in never smokers). Interestingly, our study showed a higher frequency of ipsilateral mediastinal or subcarinal lymph node metastasis (N2) in ALK-rearranged tumors compared with non-rearranged, EGFR mutated and KRAS mutated tumors, while no significant differences were found between ALK-rearranged and non-rearranged/EGFR mutated/KRAS mutated tumors with regard to N0, N1 and N3 stages. In addition, ALK rearranged lung adenocarcinomas were more frequently pT1 tumors compared with ALK-non rearranged lung cancer. In a previous study, evaluating surgically resected stage I-III NSCLCs, Paik et al. [45] found that ALK FISH-positive NSCLC cases showed lower tumor stage (pT1), but had more frequently lymph node metastases compared with ALK FISH-negative NSCLC cases. The authors suggested that ALK-rearranged lung cancer might have unique biological features with a tendency to early lymph node metastasis despite the small primary tumor size, which could explain higher incidences of ALK rearrangement in advanced NSCLC compared with surgically resectable lung cancer [45].

The frequency of BRAF mutations in the current study seems to be among the higher previously reported mutation rates [46,47,48], but is still lower than the mutation rate reported by Illei et al. [26] (6.3%), who analyzed 1006 lung cancers with NGS. Other targetable genomic alterations in NSCLC, including RET and ROS1 rearrangements and ERBB2, MET exon 14 skipping and PIK3CA mutations, are present only in a small percentage of NSCLC patients (~ 1–2% [49], ~ 2% [50], 2–4% [51,52,53], 3–4% [54,55,56,57] and 2–5% [26, 58, 59], respectively). While our study with a limited sample size of RET and ROS1 rearranged lung cancer showed no significant differences between RET or ROS1 rearranged and non-rearranged tumors regarding clinicopathological characteristics, previous investigations have reported a higher incidence of RET and ROS1 rearrangements in younger age group and never smokers [60, 61] as well as a significant association of RET rearranged NSCLC with small primary tumor size and lymph node involvement [60, 62]. According to previous reports [63], ERBB2 mutations in NSCLC are more common in females, Asian cohorts and never-smokers. While our study showed no significant association of ERBB2 mutation with female gender, we could confirm the higher prevalence of ERBB2 mutations in never smoking patients. PIK3CA mutations are more commonly encountered in squamous NSCLC [58, 59, 64] and seem to confer inferior prognosis in lung adenocarcinoma [65]. Interestingly, PIK3CA mutations have been reported to occur in parallel with other oncogenic driver mutations [66, 67], as was the case in 5 of 7 PIK3CA mutated tumors in the present study. Regarding the clinicopathological characteristics of MET exon 14 skipping mutation-positive tumors, three retrospective studies showed that MET exon 14 skipping positivity in NSCLC patients is significantly associated with advanced age [68,69,70]. In the current study, we found no significant difference in mean age between patients with and those without MET exon 14 mutated tumors. However, we acknowledge that the sample size of MET exon 14 mutated tumors was too small to draw meaningful conclusions.

Conclusion

Our study presents data on the frequency of oncogenic driver mutations in a Northeastern Swiss population with stage I-IV lung adenocarcinoma using NGS as testing method in the majority of cases. A number of studies already analyzed oncogenic driver mutation frequencies, notably EGFR, KRAS and ALK mutation rates, in different populations from European countries. However, based on the available data, the true prevalence of mutations in lung adenocarcinoma is often difficult to determine due to patient selection bias, different testing platforms used for analysis and the histological heterogeneity of tumors included in the studies. Although we cannot exclude some selection toward patients with higher likelihood of mutated tumors in the current study, a major selection bias is unlikely to have occurred because the epidemiological characteristics of our study population are similar to those of the INSIGHT study and other previous investigations. We found a relatively high EGFR mutation rate, while KRAS, BRAF, ERBB2, MET and PIK3CA mutation and ALK, RET and ROS1 rearrangement frequencies were similar to those of previous reports. EGFR and KRAS mutation was significantly associated with gender and smoking status, while ALK rearrangement was significantly associated with smoking status alone.