A meta-analysis of adjuvant EGFR-TKIs for patients with EGFR mutation of resected non-small cell lung cancer

1. Introduction

Lung cancer is one of the most common cancers in the world, with the highest mortality rate (18.0%). NSCLC (non-small cell lung cancer) accounts for 80% of all lung cancers. 5-year survival after radical resection for NSCLC patients in stages IB-IIIA remains 26% to 62%.[1] Postoperative adjuvant therapy is therefore essential. From 2003 to 2008, several large randomized controlled trials (RCT) were conducted to determine whether adjuvant chemotherapy (ACT) after radical surgery effectively improves long-term survival in this patient population.[2,3] The studies also showed that in stage IB-IIIA NSCLC cases undergoing radical surgery, cisplatin-based chemotherapy could only increase 5-year survival by 5% (40–45%).[4]

Adjuvant therapy after radical surgery has become a rational approach to lowering the risk of recurrence and improving overall survival (OS) outcomes. According to some studies, adjuvant epidermal growth factor receptor-tyrosine kinase inhibitor (EGFR-TKIs) can significantly improve the survival of NSCLC patients with EGFR mutations after radical resection compared to standard chemotherapy.[5–9] In the studies CTONG1104 and EVIDENCE, disease-free survival (DFS) was improved when Adjuvant EGFR-TKIs were compared with standard of care chemotherapy in patients with NSCLC and EGFR mutation, with hazard ratio (HR) = 0.60, 95% CI: 0.42, 0.87; Phet = 0.0054) and HR = 0.36 [95% CI: 0.24, 0.55]; Phet < 0.0001),[7,10]respectively. Furthermore, the recently updated ADAURA study found that, regardless of disease stage, patients who received osimertinib after radical surgery had a better DFS than those who received adjuvant chemotherapy (HR = 0.16, 95% confidence interval: 0.10, 0.26).[11] While those studies did not show a significant difference in OS due to DFS advantage. Simultaneously, some studies yielded negative results.[12,13] The IMPACT study found no statistically significant differences in DFS or OS, with HR = 0.92 (95% CI: 0.67, 1.28) and HR = 1.03 (95% CI: 0.65, 1.65), respectively, in patients with completely resected pathologic stage II-III non-small-cell lung cancer harboring EGFR mutations receiving gefifitinib.[13]Therefore, the use of EGFR-TKI after surgery is still debatable.

Because recent studies have reached conflicting conclusions.[5–13] We conducted this meta-analysis to compare adjuvant EGFR-TKI with conventional chemotherapy in NSCLC cases undergoing radical surgery, with the goal of determining the best treatment for patients with stage IB-IIIA EGFR mutations.

2. Methods 2.1. Strategy for literature search

This study systemically searched Embase, PubMed, Cochrane, and Web of Science databases for identifying related studies that compared EGFR-TKI with chemotherapy among NSCLC cases harboring EGFR mutation and received radical surgery from inception to March 10, 2022. The keywords included non-small cell lung cancer, EGFR, TKI, postoperative, and chemotherapy. The search terms included (“Non-small cell lung cancer” or “lung tumor” or “lung cancer”) and (“EGFR-TKI” or “EGFR-tyrosine kinase inhibitor” or “erlotinib” or “gefitinib” or “afatinib” or “icotinib” “neratinib” or “vandetanib” or “dacomitinib” or “osimertinib” or “canertinib” and (“Adjuvant” or “auxiliary” or “accessory” OR “adjunct” or “intercalated” or “alternative”).

2.2. Study selection

Study inclusion criteria were shown below:

Studies, including adult cases with the diagnosis of NSCLC (pathological stage IB-IIIA) who could receive ACT; Those evaluating the effect of EGFR-TKIs-ACT compared with placebo or chemotherapy, or TKIs-ACT compared with chemotherapy; Those that reported one or more related clinical outcomes like overall survival (OS) and DFS and had available long-time follow-up data; and Those with adequate raw data to calculate HRs and P-values. All the enrolled articles were published in English, and the publication type was not restricted.

Studies conforming to the following criteria were excluded:

Single-arm articles that reported outcomes of EGFR-TKI-ACT; Articles that had inadequate data to carry out statistical analysis; Duplicate studies; No available full texts in the original studies. 2.3. Outcomes and data extraction

Basic information from all the enrolled articles was collected by two researchers (CW and XYL). Any disagreements between them regarding study screening and data collection were settled down through mutual negotiation or by the opinion of a third researcher (RC). Furthermore, we recorded the available data like first author, publication year, case numbers and baseline features, clinical stage, tumor histology, interventions, EGFR status, outcomes, study design and phase, OS, and DFS for comparing the benefits of EGFR-TKI-ACT and traditional chemotherapy for NSCLC cases receiving radical surgery. Those original and acquired data were imported into the standard tables.

2.4. Quality assessment

The bias risk approach (Cochrane Handbook for Systematic Reviews of Interventions) was implemented by two reviewers for the independent assessment of study quality.[14] This meta-analysis also assessed the generation of sequences, concealment of allocation, missing data, blinding, selective reporting, as well as additional biased sources. Any disagreement was settled down through mutual negotiation or the opinion of a third investigator.

This work utilized the Newcastle–Ottawa Scale (NOS) in non-RCTs, including three categories, selection, outcome, and comparability. Our enrolled RCTs quality was evaluated following Cochrane Collaboration’s approach to evaluating bias risk (5.3.0) by the methodological items below, generation of random sequences, concealment of allocation, outcome assessment blinding, participant and personnel blinding, selective reporting, insufficient outcome data, as well as additional possible bias sources. The items were categorized into low, high, or unclear risk, and together they determined the general quality. Figures 1 and 2 display the risk-of-bias graph and summary. Table 1 display the Newcastle–Ottawa scale for quality assessment of non-randomized cohort studies. The opinion of a third researcher settled down disagreements.

Table 1 - Newcastle–Ottawa quality assessment scale. Study Selection Comparability Exposure 1 2 3 4 1 2 3 Angelo 2012 b a a b ab a a a Yelena 2011 b a a a ab a a b lv2015 b a a b a a a a Pan2021 b a a a ab a a b
F1Figure 1.:

Study screening flowchart. EGFR = epidermal growth factor receptor, TKIs = tyrosine kinase inhibitors.

F2Figure 2.:

Forest plots showing HR regarding DFS for adjuvant EGFR-TKI compared with placebo among NSCLC cases receiving radical surgery. CI = confidence interval, DFS = disease-free survival, EGFR = epidermal growth factor receptor, HR = hazard ratio, NSCLC = non-small-cell lung cancer, SE = standard error, TKI = tyrosine kinase inhibitor.

2.5. Statistical analysis

This meta-analysis was carried out by integrating survival data from articles conducted by Parmar and Tierney.[15,16] Log (HR) and standard error data of enrolled RCTs were collected to analyze the time-to-event data.

The I2 test was applied to analyze heterogeneity, where I2 < 50% and P > .1 indicated no heterogeneity, and the fixed-effects model was applied, whereas I2 > 50% stood for significant heterogeneity, and the random-effects model should thereby be utilized. P < .05 denotes statistical significance. Review Manager Software, version 5.3 (Cochrane Collaboration, Oxford, UK), was employed for statistical analysis.

DFS was considered the primary endpoint, and it characterized the duration between randomization and disease recurrence or death. OS was regarded as the secondary endpoint.

3. Results

Figure 3 represents the study screening flowchart. Among those 4135 studies obtained from the literature review, just 12 articles were qualified for our meta-analysis. Of which eight were RCTs, and four were RCSs. Table 2 displays enrolled study features.

Table 2 - Major features of qualified studies enrolled in this meta-analysis. Study Intervention Size EGFR mutation (%) TKI vs Control arm number Stage Median follow up (mo) Design Primary endpoint IB II III CTONG1104 2018[7] gefitinib 222 100% 111 0 74 148 36.5 RCT DFS/OS chemotherapy 111 EVIDENCE 2021[10] icotinib 322 100% 161 88 14 181 25 RCT DFS chemotherapy 161 IMPACT2021[23] gefitinib 232 100% 116 74 9 144 70 RCT DFS/OS chemotherapy 116 EVAN2018[8] erlotinib 102 100% 51 0 0 102 33 RCT DFS/OS chemotherapy 51 ADAURA2017[11] osimertinib + chemotherapy 682 100% 203 216 231 235 22 RCT DFS placebo + chemotherapy 207 osimertinib 100% 135 Placebo 136 lv2015[5] erlotinib\gefitinib\icotinib 257 138 (53.6%) 30 126 48 83 31 RCS DFS/OS Placebo 30 Angelo 2012[26] erlotinib or gefitinib 1118 284 (25.4%) 84 718 166 167 27 RCS DFS/OS placebo 202 BR19 2013[27] gefitinib 503 15 (4%) 7 260 175 67 56.4 RCT DFS/OS placebo 8 Yelena 2011[28] erlotinib or gefitinib + chemotherapy 167 100% 56 117 25 25 20 RCS DFS/OS placebo + chemotherapy 111 Li 2014[25] gefitinib + chemotherapy 60 100% 30 0 0 60 30.6 RCT DFS/OS placebo + chemotherapy 30 Feng2015[29] icotinib + chemotherapy 39 100% 21 17 10 12 46.4 RCT DFS chemotherapy 18 Pan 2021[20] icotinib 43 100% 22 0 43 0 35.5 RCS DFS/OS chemotherapy 21

DFS = disease-free survival, OS = overall survival, RCTs = randomized controlled trials.


F3Figure 3.:

Forest plots showing HR regarding OS for adjuvant EGFR-TKI compared with placebo among NSCLC cases receiving radical surgery. CI = confidence interval, EGFR = epidermal growth factor receptor, HR = hazard ratio, OS = overall survival, NSCLC = non-small-cell lung cancer, SE = standard error, TKI = tyrosine kinase inhibitor.

This work enrolled 1694 cases for meta-analysis altogether; among them, 926 received adjuvant EGFR-TKI, whereas 768 received placebo. In four studies (5, 25, 23, 18), not all patients have EGFR mutation, but these studies analyzed the data from cases harboring EGFR-mutation, in which the analyzed got positive results. All the patients included in the other eight studies had EGFR mutations.

3.1. Effects of adjuvant TKIs versus placebo or chemotherapy on DFS and OS

In eight RCTs and four RCSs, DFS was analyzed. As a result, EGFR-TKI-ACT improved patient DFS (HR, 0.47; 95% CI: 0.40, 0.55) (Fig. 4). There was obvious heterogeneity among enrolled articles by using the random-effects model (Phet < 0.00001, I2 = 80%). Also, our analysis showed that the OS (HR,0.74; 95% CI: 0.58, 0.95) (Fig. 5) after adjuvant EGFR-TKIs was better than chemotherapy, in which five RCTs (7, 8, 10, 11,23) and four RCSs (5, 20, 26, 28) were included. While there was obvious heterogeneity among those involved articles (Phet < 0.00001, I2 = 80%), the significant heterogeneity mainly comes from the study ADAURA2017 according to the sensitivity analyses. No significant publication bias was found.

F4Figure 4.:

Forest plots showing HR regarding DFS for adjuvant EGFR-TKI compared with chemotherapy among NSCLC cases receiving radical surgery. CI = confidence interval, DFS = disease-free survival, EGFR = epidermal growth factor receptor, HR = hazard ratio, NSCLC = non-small-cell lung cancer, SE = standard error, TKI = tyrosine kinase inhibitor.

F5Figure 5.:

Forest plots showing HR regarding OS for adjuvant EGFR-TKI compared with chemotherapy among NSCLC receiving radical surgery. CI = confidence interval, EGFR = epidermal growth factor receptor, HR = hazard ratio, OS = overall survival, NSCLC = non-small-cell lung cancer, SE = standard error, TKI = tyrosine kinase inhibitor.

3.2. Effects of adjuvant TKIs versus chemotherapy on DFS and OS

Four RCTs (7, 8, 10, and 23) and one RCS (20) were included in this analysis to assess DFS and OS. Adjuvant TKIs significantly improved DFS (HR,0.43; 95% CI: 0.26, 0.72) (Fig. 6) among EGFR-mutation cases. Significant heterogeneity was noted (Phet < 0.00001, I2 = 85%). At the same time, there is no significant beneficial effect of TKI treatment on OS (HR,0.79; 95% CI: 0.59, 1.05) (Fig. 7) compared with chemotherapy in the patients with NSCLC. The heterogeneity mostly comes from study Pan2021 with no significant publication bias.

F6Figure 6.:

Forest plots showing HR regarding DFS for adjuvant EGFR-TKI in combination with chemotherapy compared with chemotherapy among NSCLC cases receiving radical surgery. CI = confidence interval, DFS = disease-free survival, EGFR = epidermal growth factor receptor, HR = hazard ratio, NSCLC = non-small-cell lung cancer, SE = standard error, TKI = tyrosine kinase inhibitor.

F7Figure 7.:

Forest plots showing HR regarding OS for adjuvant EGFR-TKI in combination with chemotherapy compared with chemotherapy among NSCLC cases receiving radical surgery. CI = confidence interval, EGFR = epidermal growth factor receptor, HR = hazard ratio, OS = overall survival, NSCLC = non-small-cell lung cancer, SE = standard error, TKI = tyrosine kinase inhibitor.

3.3. Effects of adjuvant TKIs plus chemotherapy versus chemotherapy alone on DFS and OS

The analysis included three RCTs (11, 25, 29) and one RCS (26), showing that TKIs plus chemotherapy were superior to those of chemotherapy in both DFS (HR, 0.29; 95% CI: 0.15, 0.58) (Fig. 8) as well as OS (HR, 0.51; 95% CI: 0.25, 1.04) (Fig. 9). Upon sensitivity analysis, the combined results were not significantly affected by any study, despite the apparent heterogeneity (Phet = 0.001, I2 = 79%) regarding DFS. And no significant publication bias was found.

F8Figure 8.:

Risk of bias graph.

F9Figure 9.:

Risk of bias summary.

3.4. Ongoing clinical trials

According to Table 3, seven ongoing RCTs enrolling 1819 cases were conducted using the intervention model, including 2 (NCT02264210, NCT05120349) that collected early cases; the others involved the treatment of patients at different stages.

Table 3 - Patient features among ongoing research. Study (ClinicalTrials.gov) Study type Estimated enrollment Allocation Intervention mode Stage Actual study start date Estimated primary completion date NCT02193282 Interventional 450 Randomized Parallel assignment IB-IIIA 2014/7/17 2026/10/10 NCT02264210 Interventional 128 Randomized Parallel assignment IB 2014/10/15 2025/12/1 NCT02448797 Interventional 320 Randomized Parallel assignment II-IIIA 2015/5/19 2023/12/1 NCT04351555 Interventional 328 Randomized Parallel assignment II - IIIB N2 2020/4/17 2029/3/29 NCT05132985 Interventional 45 Randomized Parallel assignment II-IIIB N2 2021/11/24 2028/1/1 NCT04455594 Interventional 168 Randomized Parallel assignment IIIA-N2 2020/7/2 2025/10/1 NCT05120349 Interventional 380 Randomized Parallel assignment IA2-IA3 2021/11/15 2032/11/2
4. Discussion

The continuous innovation and research progress of EGFR-TKIs in recent years has provided clinical researchers with novel therapies and application ideas. Although TKIs have increased efficacy in EGFR mutation advanced NSCLC compared with chemotherapy,[17] whether TKI should be used in NSCLC cases receiving radical surgery remains controversial. According to our findings, adjuvant EGFR-TKIs enhanced DFS among cases with EGFR mutation after radical resection compared with chemotherapy, whereas OS was not significant. The adjuvant EGFR-TKIs plus chemotherapy demonstrated a significant beneficial effect on DFS and OS. The meta-analyze may support these results, showing that TKIs plus chemotherapy, the first-line therapy, significantly increase ORR while improving OS and DFS for advanced NSCLC cases with EGFR mutation.[18]

EGFR-TKIs have been recommended as first-line therapies for patients with EGFR-mutant NSCLC, and patients tend to benefit from adjuvant EGFR-TKI treatment in terms of DFS and OS. The ability of adjuvant EGFR-TKI treatment, however, remains unsatisfactory. According to some studies[7,8,23] EGFR-TKI DFS is superior to chemotherapy in EGFR mutation NSCLC patients. However, TKI cannot effectively prolong patients’ survival when compared to chemotherapy.[7,8,23] The recent IMPACT study found that gefifitinib as postoperative adjuvant therapy for patients did not improve DFS or OS.[10] The reason why TKI can only prolong DFS but not show DFS advantage translate to a significant OS difference in those studies (even in IMPACT both DFS and OS were not positive) may be that previous studies’ follow-up time (The median follow-up time for most experiments was no more than 40 months) were shorter than IMPACT (70 months), and the performance of TKI’s DFS was more superior in the early stage, while the DFS of TKI gradually decreased to the same level as the placebo.

Chemotherapy remains the guideline-recommended adjuvant treatment, despite having limited benefits for patients. EGFR-TKIs may inhibit sensitive mutant cancer cell growth, while chemotherapy may eliminate tumor cells and prevent micrometastasis.[22,23] Noronha et al recently discovered that TKI combined with chemotherapy can significantly improve DFS and OS in resected NSCLCs.[24,25] Furthermore, the percentage of patients receiving EGFR-TKI combined with chemotherapy who were alive and disease-free at 24 months was 89% (95% CI: 95, 99), while the placebo arm had only 58% (95% CI: 80, 89) in the ADAURA[11] study. A subgroup analysis was also performed in this study for patients after resection who received TKI combined with chemotherapy versus those who received chemotherapy alone, and DFS improved significantly, but OS did not. However, due to a lack of studies and patients, and in the study Yelena et al hold, not all of the cases they collected received chemotherapy. According to the published ADAURA results, the osimertinib group had an OS rate of 98% in 2 years, while the placebo arm had an OS rate of 85%. Therefore, TKI combined with chemotherapy may be a promising treatment for advanced NSCLC patients with EGFR mutations after radical resection.

DFS showed great heterogeneity mainly due to the different regimens, individual features, prior treatments, as well as inadequate case numbers. There were no uniform doses of EGFR-TKI or chemotherapy in the enrolled articles. Additionally, most studies did not analyze subtypes of EGFR-mutation. Moreover, the different ratios of patients of clinical tumor stages made it impossible to evaluate the risk and efficacy accurately. After sensitivity analysis, those adjusted results strengthened the study atypism as well as the risk of integrated results.

The present study has certain limitations. Firstly, four retrospective studies were comprised in this analysis which may improve the heterogeneity. Secondly, EGFR-TKIs achieve diverse median durations among diverse studies; thus, possibly leading to a certain selection bias. Moreover, due to the different sample sizes, publication bias exists. Although this analysis shows that the adjuvant EGFR-TKIs combined chemotherapy group improved DFS among cases having EGFR mutation following radical surgery, there are still numerous unanswered questions. Which stage of the patient can gain the most benefits from EGFR-TKI-ACT, and what is the preferred EGFR-TKI-ACT duration? The research included in this study failed to answer this question effectively, and well-designed experiments are needed to explore it.

5. Conclusion

In NSCLC patients with EGFR mutations who underwent radical surgery, adjuvant EGFR-TKI combined with chemotherapy showed superior effect over chemotherapy or placebo with reference to DFS and may prolong the OS up to some extent.

Author contributions

RC and OJ design the study. RC and CW analyzed the data. XYL prepared the methodology. CW and XYL carried out the formal analysis; provide the software. RC and OJ wrote the original draft. RC and XYL were responsible for language revisions. RC reviewed the edited the final manuscript.

Conceptualization: Ran Cui, Ou Jiang.

Data curation: Ran Cui, Chun Wei.

Formal analysis: Chun Wei, Xianyi Li.

Investigation: Ran Cui.

Methodology: Xianyi Li.

Software: Chun Wei, Xianyi Li.

Writing – original draft: Ran Cui, Ou Jiang.

Writing – review & editing: Ran Cui, Xianyi Li.

References [1]. Adjuvant vinorelbine plus cisplatin versus observation in patients with completely resected stage IB-IIIA non-small cell lung cancer (Adjuvant Navelbine International Trialist Association [ANITA]): a randomised controlled trial. Lancet Oncol. 2006;7:719–27. [2]. Winton T, Livingston R, Johnson D, et al. Vinorelbine plus cisplatin vs. Observation in resected non-small-cell lung cancer. N Engl J Med. 2005;352:2589–97. [3]. Pignon JP, Tribodet H, Scagliotti GV, et al. Lung adjuvant cisplatin evaluation: A pooled analysis by the lace collaborative group. J Clin Oncol. 2008;26:3552–9.

留言 (0)

沒有登入
gif