Efficacy of Crizotinib, Ceritinib, and Alectinib in ALK-Positive Non-Small Cell Lung Cancer Treatment: A Meta-Analysis of Clinical Trials

This study aimed to evaluate the efficacy of anaplastic lymphoma kinase (ALK)-inhibitors in the treatment of ALK-positive non-small cell lung cancer (NSCLC) by using a meta-analysis of clinical trials. We searched PubMed, EMBASE, Cochrane Library, and Clinicaltrials.gov by using keywords related to the topic in August 2018. The pooled effect sizes were calculated based on a random-effects model. We also performed subgroup meta-analysis by types of ALK inhibitors (crizotinib, ceritinib, and alectinib). A total of 20 clinical trials with 10 single-arm trials and 10 double-arm trials were included in the final meta-analysis. The median overall survival (OS), progression-free survival (PFS), overall response rate (ORR), disease control rate (DCR), 1 year survival rate, and 2 year survival rate were 19.14 months, 8.47 months, 62%, 78%, 74%, and 62%, respectively. ALK inhibitors showed a significantly superior efficacy compared with chemotherapy (hazard ratio (HR) for OS, 0.83; HR for PFS, 0.43; rate difference (RD) for ORR, 0.23; and RD for DCR, 0.10). The current meta-analysis of clinical trials showed the significant efficacy of ALK inhibitors in the treatment of ALK-positive NSCLC. Further head-to-head trials are needed to compare their efficacy with other types of NSCLC treatment regimens. PROSPERO registration: CRD42018085987.


Introduction
Non-small cell lung cancer (NSCLC) accounts for approximately 85-90% of lung cancers, which are the most common fatal malignancy and leading cause of cancer mortality worldwide [1,2]. It is reported that the median overall survival (OS) with platinum-based chemotherapy is approximately 7.5-28.2 months among advanced NSCLC patients, and the median progression-free survival (PFS) is approximately 2.1-6.9 months [3]. In the last decade, the treatment of advanced NSCLC has shifted into determining molecular subtypes of the disease based on oncogenic drivers, which has led to the introduction of several newly approved biological agents [4].

Risk of Bias for Randomized, Double-Blind, Placebo-Controlled Trials
Begg's funnel plot and Egger's test showed no evidence for publication bias (p > 0.05 for PFS, overall response rate (ORR), disease control rate (DCR), and 1 year survival rate; Figure 2).

Risk of Bias for Randomized, Double-Blind, Placebo-Controlled Trials
Begg's funnel plot and Egger's test showed no evidence for publication bias (p > 0.05 for PFS, overall response rate (ORR), disease control rate (DCR), and 1 year survival rate; Figure 2).    Table 3 shows the efficacy of ALK inhibitors in patients with ALK-positive NSCLC in the subgroup meta-analysis type of ALK inhibitors for each outcome in single-arm or double-arm trials. Overall, ceritinib showed shorter OS and PFS and lower ORR and DCR, compared with crizotinib and alectinib.

Summary of Findings
In the current meta-analysis of clinical trials, we demonstrated that the median OS, PFS, ORR, DCR, 1 year survival rate, and 2 year survival rate for ALK inhibitors including crizotinib, ceritinib, and alectinib in the treatment of ALK-positive NSCLC was 19.14 months, 8.47 months, 62%, 78%, 74%, and 62%, respectively. In the subgroup analysis by type of ALK inhibitor, overall ceritinib showed shorter OS and PFS and lower ORR and DCR compared with crizotinib and alectinib. As compared with chemotherapy, ALK inhibitors showed superior efficacy in the treatment of ALK-positive NSCLC.

Comparison with Previous Studies
Our findings are consistent with those from retrospective chart review studies. El. Din et al. reported that crizotinib showed a 1 year survival rate of 71.2% and an objective response rate of 70.9% [34]. Bendaly et al. reported that the ORR for ceritinib was 69% and median PFS was 12.9 months [6]. In a large, multi-country medical chart review (n = 1471) with seven countries, there was a significant improvement in complete response (odds ratio (OR), 2.65; 95% CI, 1.69-4.15) and a significant reduction in recurrence/progression (OR, 0.38; 95% CI, 0.24-0.59) [35].

Possible Mechanisms
There are several mechanisms regarding the therapeutic efficacy of ALK inhibitors in the treatment of ALK-positive NSCLC. In ALK-positive NSCLC, ALK-echinoderm microtubule-associated protein-like 4 (EML4) fusion protein activates the RAS/mitogen-activated protein kinase (MAPK), phosphatidylinositol 3-kinase (PI3K)/AKT, and janus kinase (JAK)/signal transducer, and the activator of transcription 3 (STAT) pathways are reported to play the important role in the development of NSCLC [5,36]. In the meantime, it has been shown that ALK inhibitors can bind the ALK protein to prevent the activation of NSCLC pathways [5,36].
Regarding the resistance to crizotinib, several biological mechanisms have been proposed [37]. Resistance may arise through the pathway of selective copy number gain or gene amplification (ALK-dependent) [37,38]. However, about 70% of crizotinib resistance is attributable to the abnormal activation of alternative signaling pathways involving ALK-independent growth, not to identifiable secondary resistance mutations or ALK copy number alterations [39,40]. Even though crizotinib significantly improved the treatment response compared with conventional chemotherapy, the disease progression in the central nervous system has still frequently occurred [41]. Thus, second-generation ALK inhibitors including ceritinib and alectinib with higher selectivity were designed to overcome resistance issues related to crizotinib and improve the activity of treatment therapy in the central nervous system [42].

Strengths and Limitations
To the best of our knowledge, this is the first meta-analysis to investigate the overall profile of ALK inhibitors' efficacy in the treatment of ALK-positive NSCLC. We also estimated the efficacy of crizotinib, ceritinib, and alectinib compared with chemotherapy, respectively.
However, there are several limitations in the current study. First, although we found that ALK inhibitors improved PFS in considerable trials, OS was not sufficiently evaluated due to a relatively small number of trials. As a surrogate, PFS can be evaluated faster, with fewer patients. However, OS is still considered to be the gold standard in clinical trials of oncology drugs [43]. Second, substantial heterogeneity was observed in the meta-analysis of single-arm studies for all the outcomes and double-arm studies for PFS, ORR, and DCR outcomes (I 2 > 50%). Last, due to a small number of trials, we were unable to conduct the head-to-head comparisons among different types of ALK inhibitors.

Materials and Methods
The protocol for this systematic review and meta-analysis was registered with the International Prospective Register of Systematic Reviews (PROSPERO registration number: CRD 42018085987).

Literature Search
We searched Pubmed, EMBASE, Cochrane library, and Clinicaltrials.gov databases from their inception until August 2018, limiting it to human subjects and clinical trials. The keywords for the literature search were as follows: 'crizotinib', 'ceritinib', 'alectinib', and 'ALK inhibitor' for intervention factors; 'non-small cell lung cancer' for outcome factor; 'clinical trial', 'randomized controlled trial' for study type. The bibliographies of relevant articles were also reviewed to identify additional studies. The format (abstract or full text) and language of publications were not restricted.

Study Selection and Eligibility Criteria
We included clinical trials that investigated the efficacy of three ALK inhibitors (crizotinib 250 mg bid, ceritinib 750 mg·qd, or alectinib 300/600 mg·bid) in ALK-positive NSCLC patients and reported findings on at least one of the following outcomes, i.e., OS (time from randomization to death), PFS (time from randomization to disease progression), ORR (complete response plus partial response), DCR (complete response plus partial response and stable disease), 1 year survival rate, and 2 year survival rate. The longer survival time in OS and PFS, or higher percentage of ORR, DCR, 1 year survival rate and 2 year survival rate a treatment has, the higher its efficacy. For studies using the same data, completely duplicated records were excluded, while partially duplicated records were combined to obtain the full information. Based on the eligibility criteria, two investigators (Hoang and Myung) independently selected studies to be included in the analysis.

Statistical Analyses
We used OS, PFS, ORR, DCR, 1 year survival rate, and 2 year survival rate with 95% CIs from individual studies to calculate the pooled effect time/ effect rate. For studies using the comparison group, we calculated a pooled HR for OS, PFS and a RD for ORR, DCR, 1 year survival rate, and 2 year survival rate between the two groups.
In order to measure heterogeneity across studies, we used Higgins I 2 , which estimates the percentage of total variation across studies. Negative values of I 2 are set to zero; I 2 ranges between 0% (no observed heterogeneity) and 100% (maximal heterogeneity) [39]. A random-effects model was used to calculate a pooled effect size [40]. Publication bias where 10 or more studies were available was examined by using Begg's funnel plot and Egger's test [41,44]. We also estimated the risk of bias for eligible studies based on the Cochrane Risk of Bias Tool [39]. We used the Stata SE version 14.0 software (StataCorp, College Station, Texas, USA) for the statistical analysis.

Conclusions
In conclusion, the current meta-analysis of clinical trials suggests the superior efficacy of ALK inhibitors including crizotinib, ceritinib, and alectinib in the treatment of ALK-positive NSCLC, compared with chemotherapy. Further randomized controlled trials are needed to evaluate the efficacy of different types of ALK inhibitors in head-to-head trials and the efficacy of those compared with other types of NSCLC treatment regimens.

Conflicts of Interest:
The authors declare no conflict of interest.