Background
Lung cancer is the leading cause of cancer-related mortality worldwide, and non-small cell lung cancer (NSCLC) is the most common form of lung cancer. Many NSCLC patients present with an advanced disease stage upon initial diagnosis [
1]. Patients with tumors that harbor activating mutations in the epidermal growth factor receptor (EGFR) benefit greatly from treatment with EGFR tyrosine kinase inhibitors (TKIs) compared with patients whose tumors lack these mutations [
2‐
7]. One such EGFR-TKI is the orally administered, targeted agent erlotinib, which inhibits the tyrosine kinase domain of the EGFR. Erlotinib is approved for second-line use based on the positive results of a phase 3 BR.21 trial [
8] in which erlotinib improved overall survival (OS) compared with the best supportive care. Erlotinib also has clinical benefits as a first-line therapy for advanced NSCLC. The tumor response rate was 10-20%, and the median survival duration was 10.9-12.9 months in phase 2 studies [
9,
10]. However, almost all patients suffered from tumor progression and inevitably became resistant to EGFR-TKIs within 8-12 months (a phenomenon referred to as acquired resistance).
Currently, the standard method of measuring the efficacy of a lung cancer treatment is anatomical imaging, including computed tomography (CT) scanning, which measures the size of malignant lesions before and after treatment. However, the use of anatomical imaging methods to assess treatment efficacy presents a number of drawbacks, the most critical of which is a delay in treatment due to changes in tumor size. Furthermore, the efficacy of targeted drugs is often not reflected by a change in tumor size but rather by changes in cell metabolism. Therefore, the identification of clinical biomarkers in patients with EGFR mutations may prove useful when anatomical analyses are not feasible.
Several serum markers are considered to be prognostic and predictive markers of NSCLC. Among these markers, carcinoembryonic antigen (CEA) is a sensitive and useful tumor marker for cancer diagnosis and prognosis and the assessment of therapy [
6‐
8]. According to recent reports, CEA is a significant predictor of sensitivity and survival in patients treated with gefitinib [
9‐
11]. The present study (1) compared the significance of CEA levels with other clinical characteristics (i.e., age, sex, smoking history, performance status [PS], and CYFRA1-1) and (2) determined whether the serum CEA levels correlated with EGFR-TKI resistance. This correlation would permit the use of CEA as a biomarker in NSCLC patients and would aid in identifying treatment candidates with reversible and irreversible EGFR-TKI resistance and candidates for whom an early intervention with combined chemotherapy and radiotherapy is more appropriate.
Discussion
EGFR mutation predicts the efficacy of EGFR-TKIs in patients with advanced NSCLC. However, acquiring an adequate tissue sample for an EGFR mutational analysis is not often feasible, particularly in patients with advanced disease [
2,
8,
14]. A recent study reported that molecular analyses of circulating tumor cells obtained from the peripheral blood of patients with lung cancer was useful for monitoring changes in epithelial tumor genotype during the course of treatment. However, this type of molecular analysis can be difficult due to the requirement of a specific, microfluidic-based device - the CTC chip. Moreover, there are approximately 486 types of EGFR-TKI domain mutations across 87 species, and new mutations are continually being identified [
15,
16]. Recently, the attention moved to the possibility of isolation and analysis of cell-free tumor DNA (cftDNA) that, to date, represents the best candidate for identification and monitoring of molecular tumor-related alterations in blood of patients with cancer [
17]. Circulating DNA fragments carrying tumor specific sequence alterations cftDNA are found in the cell-free fraction of blood, representing a variable and generally small fraction of the total circulating DNA. cftDNA has a high degree of specificity to detect
EGFR gene mutations in NSCLC. Fragments of circulating DNA were isolated in plasma many years ago [
18]. In particular, patients with cancers present higher levels of circulating DNA comparing to healthy volunteers because of the presence of tumoral counterpart, which express the same molecular abnormalities expressed by DNA of primitive mass [
19]. The elevate cellular turn over and consequent cellular necrosis and apoptosis cause a massive release of tumoral DNA into the bloodstream were it can be isolated and analyzed. Therefore, tumor size, localization and vascularity may influence cftDNA plasmatic levels. It is also possible that part of cftDNA comes from CTCs lysis [
19]. The analysis of cftDNA, defined as liquid biopsy, could be repeated every time needed and without any discomfort for patients. Moreover, the mutational analysis of cftDNA demonstrated a signicantly better sensitivity if compared with CTCs one, establishing cftDNA as the best circulating source for molecular analysis [
20]. Information derived from liquid biopsy could be used in future for early cancer diagnosis, assessment of genetic determinants for targeted therapies, monitoring of tumor dynamics and early evaluation of tumor response, identification of resistance mechanisms [
19]. cftDNA could be a relevant biomarker to molecular diagnosis and monitor treatment resistance, because of its sensitivity and specificity, but it really needs reproducible and standardized methods, both for the extraction and for its analyses. Regarding the mutation analysis of cftDNA, a large number of technologies is now available to analyze mutations in cftDNA, including automatic sequencing, real-time polymerase chain reaction (PCR) platforms, mass spectrometry (MS) genotyping, ampli cation protocols with magnetic beads in oil emulsions [beads, emulsion, ampli cation and magnetics (BEAMing)] and next-generation sequencing (NGS), digital PCR platforms [
21‐
25]. The sensitivity range of the available techniques varies from 15 to 0.01%, but one of the major gaps in this field is the lack of standardization of techniques, in order to understand how those techniques are cost-effective and reliable to the clinical needs.
Therefore, simpler and more accessible predictors of EGFR mutations, such as surrogate markers, are necessary. CEA is the product of the CEACAM5 gene, which is expressed only in epithelial cells. CEA is found more abundantly on the apical surface of the gastrointestinal epithelium but is also found in other mucosal epithelia cells, such as in the lung [
26]. Although CEA was often falsely elevated in smokers and in patients with restrictive or obstructive pulmonary diseases [
27‐
29], abnormally elevated CEA levels were reported in 30-70% of patients with NSCLC. Abnormally elevated CEA levels are most frequently observed in patients with adenocarcinoma and advanced stage carcinoma [
30]. In addition, high serum CEA levels are associated with a poor prognosis in patients with NSCLC, regardless of treatment [
30,
31]. According to Japanese scholars, patients with elevated serum CEA levels responded better to gefitinib. Furthermore, recurrent lung adenocarcinoma patients with high serological CEA levels have a higher EGFR mutation rate after surgery and higher serological CEA levels. These findings are attributed to a possible anti-apoptotic signal in the mutant EGFR pathway that could elevate the expression level of the CEA protein [
32]. However, the specimens used for genetic testing were surgical specimens obtained prior to disease recurrence and may not represent all the biological characteristics of a recurrent tumor [
33]. In our study, the serum CEA level in the EGFR gene mutation group was significantly higher than in the non-mutated group. Both the univariate and multivariate analyses indicated that the serum CEA levels correlated with EGFR mutations (higher serum CEA levels were associated with higher EGFR gene mutation rates). Our data are similar to the findings of Okamato et al. [
34]. Shoji et al. [
35] reported that the rate of EGFR gene mutation significantly increased as the serum CEA levels increased (for serum CEA levels <5, ≥5 (but <20), and ≥20, the rates of EGFR gene mutation were 35, 55 and 87.5%, respectively;
p = 0.040).
Several reports have described the relationship between serological markers and the curative effect of EGFR-TKIs. However, these reports did not perform EGFR mutation testing or dynamic monitoring of CEA levels to predict EGFR-TKI resistance. Therefore, these reports cannot determine the most effective treatment for early intervention. Despite the high responsiveness of tumors bearing activating EGFR mutations, almost all patients become resistant to TKIs. Multiple molecular mechanisms may underlie this resistance, including secondary EGFR mutations, bypassed signaling activation, and phenotypic transformation. Because multiple molecular mechanisms may lead to EGFR-TKI resistance, it is important to non-invasively detect tumors refractory to EGFR-TKI treatment and identify the mechanisms underlying this resistance. Thus, the therapy could be effectively tailored to each patient. Based on previous reports, the function of CEA has not been elucidated. However, as a cell surface adhesion protein, CEA may play a role in cell-cell adhesion [
36]. Overexpression of CEA is thought to play a role in tumorigenesis [
37]. Furthermore, CEA has a dominant effect in blocking differentiation, and CEA cooperates with Myc and Bcl-2 during cellular transformation [
38]. Furthermore, CEA can inhibit cell death induced by a loss of anchorage to the extracellular matrix (anoikis) [
39]. If CEA is upregulated following activation of the EGFR pathway, its serum levels may trigger an EGFR mutation. Although these findings suggest that CEA may have anti-apoptotic effects in cancer cells, a direct relationship between high CEA levels and patient responses to EGFR-TKIs has not yet been established and requires additional research.
We found that a persistently high level of CEA after treatment with a reversible EGFR-TKI can successfully identify patients with NSCLC cells that are resistant (perhaps because of the occurrence in the EGFR kinase domain of a T790 M secondary mutation that prevents EGFR-TKI binding and subsequent growth arrest). Furthermore, when the CEA level was 16.2 times greater than normal, the elevation was associated with distant metastasis (Table
2). According to Sequist et al. [
39], molecular analyses of repeated lung biopsies from these patients are needed to identify different mechanisms of acquired resistance. A potential clinical application of our observations could be the development of a test for patient responsiveness to EGFR-TKI treatment using non-invasive serum tumor markers. The information provided by this test may facilitate the selection of patients as candidates for therapy with reversible or irreversible EGFR-TKIs and the development of therapeutic strategies for overcoming resistance in patients with refractory NSCLC. Tumors with high CEA expression may possess an increased capacity to develop distant metastases (perhaps due to vascular-tumoral cell-cell adhesion processes). CEA serum levels may identify patients with a high risk of metastasis development prior to CT scans. Other cell adhesion molecular markers associated with lymph node metastasis, such as the chemokine receptors CCR7, CXCR3 and CCL21, could be related to distant metastasis development. Thus, studies of their association with distant metastasis development are justified.
In our study, the OS-associated factors were age, clinical stage, and serum CEA levels. In many neoplasms, a high serum CEA level predicts residual disease or tumor relapse in patients without normal-range serum levels after surgery [
40]. In fact, Iwasaki et al. proposed a formula to evaluate mortality risk based on CEA serum levels, histological type, and the presence of positive mediastinal lymph nodes [
41] High CEA serum levels may reflect micrometastatic disease, although we detected no differences in the CEA serum levels between patients of different clinical stages. This observation suggests that the prognostic role of high CEA serum levels may be completely accounted for by tumor change. CEA represents an important tumor marker associated with several physiopathological CEA expression is induced by hypoxia inducible factor α (HIF-α), suggesting that CEA plays an important role as a micro-environmental factor during tumorigenesis and confers a worse prognosis.
To our knowledge, the clinical assessment of lung cancer treatment uses the RECIST criteria as the gold standard for response evaluations. However, early diagnostic CT scans for response evaluations in patients receiving EGFR-TKI therapies have severe limitations. EGFR-TKI therapy is expected to induce a response via cytostasis, rather than an objective morphologic response. The RECIST criteria are further confounded by structural abnormalities, both before and after treatment, which may not actually be tumors.
The limitations of this study should be acknowledged. There is no consensus regarding the optimal timing for performing either CT scans or serum CEA measurements during or after prolonged treatments. According to RECIST version 1.1, the best radiologic response evaluation can be obtained at least 4 weeks after the initiation of therapy. In our study, we performed CT scans every 2 weeks after the initiation of therapy. Therefore, the relatively small number of patients exhibiting a radiologic response could be explained by the timing of the CT scans. In addition, normal serum CEA levels in this study ranged from 0.0 to 3.4 ng/ml, which are lower than the previously reported value of 5.0 ng/ml [
27,
42,
43].