Crizotinib with or without an EGFR-TKI in treating EGFR-mutant NSCLC patients with acquired MET amplification after failure of EGFR-TKI therapy: a multicenter retrospective study

Lung cancer is one of the most common causes of death from cancer worldwide [1]. Over the past decades, advances in molecular analysis and targeted therapies have evolved the treatment and survival rates for patients with lung cancer. Epidermal growth factor receptor (EGFR) is one of the major driving mutations in non-small cell lung cancer (NSCLC) [2].

EGFR tyrosine kinase inhibitors (TKIs) have been used as first-line treatment of advanced NSCLC patients harboring EGFR mutations, which successfully improved the response rate (RR) and prolonged survival compared with chemotherapy [3‐7]; however, with first-generation EGFR-TKIs, most patients exhibit disease progression with a median progression-free survival (mPFS) ranging from 9 to 13 months. Various mechanisms for the acquisition of resistance to EGFR-TKIs have been identified, among which approximately 50% of cases involving acquired resistance (AR) are due to a secondary T790M mutation in exon 20 of EGFR gene [8]. Other causes of AR include bypassing signaling activation, phenotype transition, and aberrant downstream signaling pathways [9‐13]. One example of bypassing signaling activation is acquired mesenchymal epithelial transition factor receptor (Met) gene amplification [14], causing c-Met amplification to circumvent the inhibited EGFR phosphorylation kinase pathway and activate downstream signal transduction, such as the PI3K/AKT pathway, thus developing resistance to EGFR-TKI and facilitating tumor cell proliferation. It has been reported that 5%–20% patients harbor c-Met amplifications with acquired resistance to EGFR-TKIs [14‐16].

As a MET inhibitor, the efficiency of crizotinib in treating NSCLC patients with MET gene mutations has been demonstrated [17, 18]. The efficacy of combining EGFR-TKIs with MET inhibitors, such as crizotinib, capmatinib, INC280, or savolitinib, has been determined [19]; however, few clinical trials with a large sample size have compared the efficacy of crizotinib monotherapy and crizotinib combined with EGFR-TKIs as a therapeutic strategy and prognosis of EGFR-mutated patients with acquired c-Met amplification. In the present multicenter retrospective study, we analyzed the characteristics of EGFR-mutant NSCLC patients with acquired MET amplifications and the outcomes of crizotinib monotherapy and crizotinib plus EGFR-TKIs after failure of previous lines of EGFR-TKI treatment.

Both crizotinib monotherapy and crizotinib plus EGFR-TKI treatment provided promising outcomes. This is the report with a relatively large sample size that evaluates the efficacy of crizotinib for the acquired MET amplification after EGFR-TKI therapy in Asian NSCLC patients.

MET amplification is one of the mechanisms that contributes to acquired resistance to EGFR-TKIs. According to previous reports, 5%–20% of patients with metastatic EGFR-mutated NSCLC develop acquired resistance to EGFR-TKIs through MET amplification [21‐23]. Previous studies have reported patients with EGFR-mutant NSCLC and acquired MET amplification treated with MET inhibitors [24‐29]. Crizotinib is an effective MET inhibitor for patients with MET amplification [17]; however, the outcomes of patients treated with crizotinib after developing resistance to EGFR-TKIs has not been determined.

In the current study we reported that the incidence of an acquired resistance mechanism due to MET amplification was higher in patients with an exon 21 L858R mutation (88.9%) than an EGFR exon 19 deletion (11.1%). Emergence of the T790M mutation is regarded as the most common mechanism of acquired resistance to EGFR-TKIs. The incidence of acquired T790M mutations differs between patients with exon 19 deletions and patients with exon 21 L858R mutations. Jenkins et al. [30] conducted T790M detection testing in the AURA (327 patients) and AURA2 trials (383 patients), which found that patients with exon 19 deletions are at a higher risk of developing T790M mutations than patients with L858R mutations (73% vs. 58%; P = 0.0002). Piotrowska et al. [31] conducted a similar study and reported that the corresponding rates of T790M mutations were 69% (94/137) and 30% (41/137), respectively. An observation trial involving a greater number of patients is expected to verify this trend.

The MET gene is a clinically relevant mutation that predicts the response to treatment of MET inhibitors. It is well-known that targeted therapy based on genetic testing improves the survival of cancer patients. In our study, four patients who received chemotherapy rather than crizotinib therapy had a significantly shorter mOS compared with patients who received crizotinib treatment (39.5 months vs, 17.0 months, P < 0.001). Hence, patients with acquired c-MET amplification may benefit from crizotinib treatment. A large sample prospective clinical study is needed for further evaluation.

In addition, the efficacy and survival of such patients treated with crizotinib monotherapy or crizotinib plus an EGFR-TKI are unclear. Met gene-mediated acquired resistance to EGFR-TKIs involves the activation of signaling pathways downstream from PI3K/mTOR [14, 32]. MET amplification is sensitive to treatment with MET inhibitors, including crizotinib or other MET-TKIs [33, 34], which supports the approach to combining EGFR-TKI with a MET inhibitor to overcome acquired resistance. Yoshimura et al. [25] reported the first case of successful crizotinib monotherapy in EGFR-mutant NSCLC that acquired MET amplification during erlotinib therapy and PR, but PFS was not observed. Previous studies have investigated the efficacy of INC280 (a c-MET inhibitor) in combination with gefitinib in treating NSCLC c-MET-positive patients with EGFR-TKI resistance, which resulted in an ORR of 29% and a DCR of 73%; the mPFS was 5.6 months [33]. It has been demonstrated that MET inhibitors in combination with EGFR-TKIs can overcome EGFR-TKI resistance mediated by aberrant c-Met activation in NSCLC preclinical models [35, 36]; however, the efficacy and survival data of such patients treated with crizotinib have rarely been published. One case report showed that the PFS for a patient treated with crizotinib plus gefitinib was 8.8 months [25]. Veggel et al. [37] reported that crizotinib treatment resulted in short-lasting responses in patients with EGFR mutation-positive NSCLC who acquired c-MET amplification after EGFR-TKI therapy (all 8 patients). Veggel et al. [37] showed a PR of 50% and a mPFS of 1.4 months. The patients were treated with crizotinib monotherapy (n = 2) or in combination with a first (n = 3) or third (n = 3) generation EGFR-TKI. In our study, the mPFS of patients treated with crizotinib plus an EGFT-TKI was longer than crizotinib monotherapy (12.6 months vs. 6.0 months, P = 0.315). We believe the efficacy of crizotinib may be involved with the co-occurrence of other gene mutations, therapeutic models, and the cMET-to-CEP7 ratio.

In our study, two patients (33.3% [2/6]) had grade 3 adverse events and received dose adjustments. There is only one phase 1 study which reported 100 mg and 150 mg twice a day as the MTD for erlotinib and crizotinib, respectively [38]. To date, no standard combination treatment regimens have been established. Therefore, the patients received a standard dose of crizotinib whether in the crizotinib alone or combination treatment group. We believed that patients who had already been treated with a EGFR-TKI might be better able to tolerate adverse reactions than patients who were initially treated with an EGFR-TKI and crizotinib combination. Veggel et al. [37] reported that a crizotinib dose modification was necessary for three patients, two of whom were treated with crizotinib monotherapy and one of whom was treated with crizotinib and osimertinib. Among the three patients who were treated with crizotinib plus erlotinib, all patients tolerated the standard dose with no dose adjustment. Therefore, patients with an acquired MET, combination therapy did not increase the frequency of adverse reactions and the patients tolerated standard doses. In addition, combination of the two drugs may increase hepatic toxicity. Icotinib was associated with a lower number of treatment-related adverse events when compared to gefitinib with respect to overall incidence and diarrhea, suggesting that icotinib might have a better safety profile. Icotinib has a better safety profile than gefitinib for other major adverse events [39]. Furthermore, high-dose icotinib (250 mg tid) has tolerable toxicity in NSCLC patients harboring 21 L858R mutations [40]. Hence, icotinib is relatively well-tolerated, and adverse reactions are controllable in patients receiving combination therapy. Of course, in clinical treatment, we must closely monitor adverse reactions and make drug dose adjustments in a timely manner.

The limitations of our study should be mentioned. Despite the fact that the present study recruited the largest patient sample compared with similar previous studies, only six of 18 patients were treated with crizotinib plus an EGFR-TKI. The power of the statistical analysis may have been compromised due to the limited group members. Furthermore, other potentially existing mutations were not considered.

EGFR-mutated NSCLC patients developed resistance to EGFR-TKIs due to MET amplification with EGFR-TKI plus c-Met inhibitors by providing simultaneous inhibition of both pathways. Patients with acquired MET amplification benefited from crizotinib monotherapy and crizotinib plus EGFR-TKI treatment. Tolerance to combination therapy was acceptable, but requires close monitoring. Randomized trials or well-designed clinical trials are needed in the future to establish the role of crizotinib in patients with acquired resistance.