Background
Lung cancer remains the leading cause of cancer deaths worldwide [
1]. Non–small-cell lung cancer (NSCLC) is the most prevalent histological type, accounting for 85% of all lung cancers [
2].
Molecular targeted therapies for specific genetic alterations such as epidermal growth factor receptor (EGFR) have significantly improved the progression-free survival (PFS) and overall survival (OS) and profoundly shifted the treatment landscape in recent decades [
3‐
5]. However, most patients who initially respond to EGFR-TKIs will inevitably develop resistance within 1 year [
6‐
8]. To prevent or delay the emergence of acquired resistance to EGFR-TKIs, combination therapy with chemotherapy or antiangiogenic agents plus EGFR-TKIs are an emerging trend and have been evaluated in several clinical trials [
9‐
12]. Bevacizumab is one of the commonly used antiangiogenic monoclonal antibodies that target the vascular endothelial growth factor (VEGF) signaling pathway [
13]. Preclinical studies suggested that simultaneous inhibition of the EGFR and VEGF/VEGFR pathways could yield a biologically synergistic effect on antitumor activity and could overcome primary and acquired EGFR-TKI resistance [
14,
15]. In the JO25567, NEJ026, and ARTEMIS-CTONG1509 trials, bevacizumab plus erlotinib showed prolonged PFS as compared with erlotinib monotherapy [
9‐
11]. Several other studies also showed that bevacizumab plus EGFR-TKIs significantly prolonged PFS with acceptable toxicity profile than EGFR-TKI monotherapy for patients with
EGFR-mutant NSCLC [
12,
16,
17].
However, data on the effect of concomitant gene mutations at baseline on the clinical efficacy of the EGFR-TKI plus bevacizumab combination therapy are still lacking. Despite treatment with EGFR-TKI, the prognosis remains poor for patients with concomitant gene mutations such as
TP53 [
18,
19]. The optimal first-line treatment for NSCLC patients with various concomitant mutations remains unclear. Our previous clinical study revealed that EGFR-TKIs plus bevacizumab was associated with a significantly higher systemic and intracranial objective response rate (ORR) and significantly longer systemic and intracranial PFS [
20]. On this foundation, we analyzed the therapeutic effect of the addition of bevacizumab to EGFR-TKI among Chinese patients with
EGFR-mutant locally advanced/advanced NSCLC. Our main research aim is to evaluate the effect of concurrent gene mutations on patient prognosis. We also constructed two nomogram prediction models for PFS and OS, respectively, based on molecular features such as the presence/absence of concurrent gene mutations, and clinical features such as the specific location of metastasis. We also explored the molecular mechanisms of acquired resistance to the combination therapy.
Discussion
Prior clinical trials have demonstrated the improvement of PFS with the addition of bevacizumab to the EGFR-TKI treatment [
9‐
12,
21]. However, a major challenge remains in the selection of patients who would derive more benefit from the addition of bevacizumab. In addition, the mechanism underlying this improvement is still largely unknown. In this exploratory study, we identified the landscape of the concurrent genomic alterations in
EGFR-mutant NSCLCs and explored the subset of patients who can benefit from the addition of bevacizumab into first-line EGFR-TKI regimen. Our study demonstrated that the addition of bevacizumab may overcome the clinical and molecular factors that negatively impact EGFR-TKI efficacy and thus enhance its therapeutic effectiveness.
In our study, the median PFS was 14.0 months in the A+T group and 10.5 months in the T group, which were shorter than the reported outcomes of the NEJ026 study (median PFS 16.9 months in erlotinib plus bevacizumab group and 13.3 months in erlotinib alone group) and the ARTEMIS-CTONG1509 (median PFS 17.9 months in erlotinib plus bevacizumab group and 11.2 months in erlotinib alone group [
10,
11]. We consider our observation to be acceptable due to the complicated situations in the real-world clinical practice. Interestingly, our study suggested a significantly longer OS in the A+T group than in the T group, which is in contrast to the lack of OS difference observed in the JO25567, NEJ026, and ARTEMIS-CTONG1509 studies [
9‐
11]. We recommend further studies with a larger sample size and longer follow-up to clarify this inconsistency.
The concomitant genetic alterations in either tumor suppressor or oncogenic driver genes, or
EGFR amplification, were frequently detected in our cohort. Among these genetic alterations, we observed that patients having concomitant mutations in the tumor suppressor genes had longer PFS in the A+T group than in the T group, indicating the favorable role of the combination treatment strategy in patients with concurrent tumor suppressor gene mutations. Previous study proved that harboring concurrent tumor suppressor gene mutation was a negative prognostic factor for EGFR-TKI treatment [
22]. Our study further demonstrated that these patients could benefit from the addition of bevacizumab, which to some extent can improve not only the PFS, but also the OS of this subset of patients.
Of the tumor suppressor gene alterations, we observed that
TP53 mutation, which frequently occurs in many human malignancies and is associated with cancer development and progression, was also the most common concurrent genetic alteration in our cohort (41%). This may contribute to the clinical significance of the addition of bevacizumab to EGFR-TKI regimen in our population, particularly among the patients with concomitant mutations in
TP53 or other tumor suppressor genes. Retrospective studies have shown that the presence of mutant
TP53 is associated with poorer PFS with EGFR-TKI therapy [
18,
19,
23‐
25]. Consistent with these studies, our study suggested that concomitant
TP53 mutation negatively impact the PFS for EGFR-TKI monotherapy; however, this was not observed in the A+T group. Besides, compared with EGFR-TKI monotherapy, A+T prolonged the PFS in
TP53-mutant patients, which was not observed among
TP53 wild-type patients. We speculated that the addition of bevacizumab counteracted the negative effect of
TP53 mutation on EGFR-TKI efficacy and
TP53 mutation may serve as a biomarker for predicting the subset of patients who may benefit from the bevacizumab combined with EGFR-TKI. This conclusion was to some extent supported by some prior studies evaluating the impact of
TP53 mutation on the efficacy of bevacizumab combined with chemotherapy in other solid tumor types including metastatic colorectal cancer and advanced endometrial cancer [
26,
27]. Studies have also reported that
TP53 mutation was positively correlated with VEGF-A expression [
28].
We constructed nomograms for predicting the risk of progression (PFS) and the risk of death (OS) of the T group based on two significant factors: the presence/absence of concurrent genetic mutations in tumor suppressor genes and oncogenic driver genes and the location of metastasis (locally pleural, brain or liver metastases). Since the models did not impact the PFS or OS for the A+T group, further analyses revealed that the factors included in the nomogram had no effect on PFS of A+T group, while the presence of brain or liver metastasis shortened their OS. This suggested that the factors that negatively impact the efficacy of EGFR-TKI treatment became insignificant when bevacizumab was added to the regimen. Hence, we concluded that the addition of bevacizumab may improve the efficacy of EGFR-TKI through overcoming the negative impact of the clinical and molecular factors which originally affect its efficacy. In the era of precision oncology, other clinical and molecular factors should be taken into account when deciding the treatment strategy. According to the nomogram model, patients with lower scores, or those without concurrent mutation or only have local/pleural metastasis could benefit from single-agent EGFR-TKI regimen. Meanwhile, patients with higher scores, or those with concurrent mutations or have distant metastasis could benefit from A+T combination therapy.
Limitations should be noted in our study. Although the propensity score matching system was used to identify the patients with similar baseline clinical characteristics between the two groups, selection bias was inevitable due to the retrospective nature of our study. In addition, the genomic mutational status was exploratory and had not been verified by another technology. To verify our results, further prospective multi-center studies are recommended.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.