Pemetrexed versus Gefitinib as Second-line Treatment for Advanced Non-small Cell Lung Cancer: A Meta-analysis Based on Randomized Controlled Trials

Abstract Objective To investigate the clinical efficacy and toxicity of Pemetrexed versus Gefitinib as second-line treatment for advanced non-small cell lung cancer (NSCLC). Methods By systematically searching the electronic databases of Pubmed, CENTRAL, Cochrane, EMBASE, ASCO, and CBM, open published randomized clinical trials (RCTs) relevant to clinical efficacy and toxicity of Pemetrexed versus Gefitinib as second-line treatment of advanced NSCLC were included in the meta-analysis. Data of objective response rate (ORR) and drug related toxicity were extracted from the original publications and pooled by random or fixed effect method. Results Fourteen clinical trials related to Pemetrexed versus Gefitinib as second-line treatment for advanced NSCLC fulfilled the inclusion criteria and were included in the meta-analysis. The pooled results show that the ORR (RR=0.81, 95% CI:0.56–1.16, p=0.25) and DCR (RR=1.11, 95% CI:0.94–1.31, p=0.24) were not statistical different for Pemetrexed versus Gefitinib as second-line treatment of advanced NSCLC. However, the pooled data demonstrated the risk of developing skin rash (RR=0.10, 95% CI:0.03–0.30, p=0.00) and diarrhea (RR=0.31, 95% CI:0.15–0.67, p=0.003) in patients with Pemetrexed was significantly lower than that of Gefitinib through random effect model analysis, but the incidence of neutropenia in Pemetrexed group was significantly higher than that of Gefitinib with statistical difference (RR=7.62, 95% CI:3.71–15.66, p=0.00). Conclusion Pemetrexed was not inferior as second-line treatment for advanced NSCLC compared to Gefitinib for tumor response. However, Pemetrexed had higher incidence of neutropenia but lower risk of developing skin rash and diarrhea.


Introduction
Lung cancer (including non-small cell lung cancer and small cell lung cancer) is the most diagnosed malignant carcinoma for males and females word-wide, and is the leading cause of cancer related death for men and the second for women [1]. Non-small cell lung cancer (NSCLC) is a main type of lung cancer, accounting for about 80% of all type lung cancers. About 75% of NSCLC were at advanced or locally advanced metastatic stage when first diagnosis and lost the opportunity for operation [2]. For these advanced stage NSCLC patients, a platinum based two-drug combination chemotherapy regimen is the first-line chemotherapy and is recommended by the National Comprehensive Cancer Network (NCCN). Compared to best supportive care, platinum based twodrug chemotherapy can improve the overall survival and progression free survival (PFS) [3,4]. However, most of the patients developed drug resistance after several cycles of platinum based first-line chemotherapy. A second-line chemotherapy was recommended for these patients with good performance status [5]. The second-line chemotherapy drugs includes Docetaxel, Pemetrexed, and small molecule tyrosine kinase receptor inhibitors (epidermal growth factor receptors) tyrosine kinase inhibitor (EGFR-TKI) Gefitinib and Erlotinib [6][7][8][9]. Pemetrexed is a new-generation antifolate agent, approved for the treatment of mesothelioma and NSCLC. Pemetrexed works by disrupting folate-dependent metabolic processes which are essential for cancer cell replication and survival. Previously published randomized clinical trials have compared the efficacy and toxicity of Pemetrexed versus Gefitinib as secondline treatment for advanced NSCLC [10][11][12]. However, the results were not conclusive because of small sample sizes with limited statistical power. Therefore, in our present work we investigate the clinical efficacy and toxicity of Pemetrexed versus Gefitinib as second-line treatment for advanced NSCLC in order to provide more information for clinical use of these two drugs.

Publication electronic searching
The randomized clinical trials (RCTs) systematic electronic searching process was performed according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement flow chart ( Figure  1). Prospective RCTs related to Pemetrexed versus Gefitinib as second-line treatment for advanced NSCLC published before January 2019 were screened in the electronic databases of Pubmed, CENTRAL, Cochrane, EMBASE, ASCO, and CBM. The electronic searching text words were: non-small cell lung cancer/NSCLC, Gefitinib, Iressa. The publication search was restricted to human beings and with the language restriction of English and Chinese.

Publication inclusion and exclusion criteria
The studies were screened and included based on publication type, patients, treatment, and outcomes. The included publications fulfilled the following requirements: (i) Prospective randomized controlled trials. (ii) The subjects included in each individual study should be limited to NSCLC with advance stage and previous chemotherapy treatment. (iii) The second-line chemotherapy restricted to Pemetrexed versus Gefitinib. (iv) The objective response rate (ORR), disease control rate (DCR), and chemotherapy related toxicity were proved in the original included studies. (v). Studies published in Chinese or English. The publication with the following features were excluded: (i) Retrospective clinical observation. (ii) Small cell lung cancer. (iii) Patients received previously chemotherapy of Pemetrexed or Gefitinib. (iv) Duplicated publication or data. Publication searching and inclusion was performed by two reviewers independently and discussed when disagreement was encounter.

Data extraction
The data was extracted by two reviewers independently. A data extraction table was prepared before extracting the data. The table includes author name, journal of the publication, publication year, dosage of Pemetrexed or Gefitinib, age of the subject, and outcomes (ORR, DCR, skin rash, diarrhea, and neutropenia).

Statistical analysis
Stata 11.0 statistical software was used to analysis the data. The dichotomous data ORR, DCR, frequency of skin rash, diarrhea, and neutropenia is shown as risk ratio (RR) when comparing Pemetrexed and Gefitinib. A RR＜1 indicates a low risk of ORR, DCR, skin rash, diarrhea, and neutropenia. Before pooling the data, we first evaluated the statistical heterogeneity through the I 2 test. If I 2 ＜50%, the statistical heterogeneity was significant and the data was pooled by random-effect method. Inversely, if there was no significant heterogeneity, the fixed-effect method was sued.

Pooled drug related toxicity
The drug related toxity inclued skin rash, diarrhea, and neutropenia. Study numbers 10, 13 and 8 reported the incidence of skin rash, diarrhea, and neutropenia respectivley. Significant statisticla heterogeniety was found in skin rash and diarrhea (p<0.05), but not neutropenia (p>0.05). The pooled results showed that the risk of developing skin rash (RR=0.10, 95% CI:0.03-0.30, p=0.00) (Figure 4) and diarrhea (RR=0.31, 95%  (Figure 5) in the Pemetrexed group was significant lower than that of the Gefitinib group, as determined through random effect model analysis. However, the incidence of neutropenia in the Pemetrexed group was significantly higher than that of the Gefitinib with statistical difference (RR=7.62, 95% CI:3.71-15.66, p=0.00), Figure 6.

Publication bias
The publication bias of Pemetrexed versus Gefitinib as second-line treatment for advanced NSCLC was assessed through Begg's funnel plot. The plot has general left and right symmetrical which indicates no significant publication bias, Figure 7.

Discussion
Epidemiology studies have demonstrated that NSCLC has become the leading cause of cancer related death wordwide. However, most of the cases were at advance stages and must receive chemotherapy as the main treatment method. Pemetrexed was the most clinical used drug for second line chemotherapy of NSCLC. In 2004, the Food and Drug Administration (FDA) of the United States approved Pemetrexed as the second-line for the treatment of locally advanced or metastatic NSCLC in patients with non-squamous histology [24,25]. Prospective randomized clinical phase III trials proved that Pemetrexed maintenance chemotherapy can improve the prognosis of patients with advance non-squamous cell NSCLC [4]. The tumor inhibiting activity of Pemetrexed works by inhibiting three enzymes used in purine and pyrimidine synthesis, thymidylate synthase (TS), dihydrofolate Dai (2013) Sun (2012) Lin (2016) Song (2015) Wang (2012) Zhang (2012) Xu (2015) Kim (    Dai (2013) Liu (2015) Kim (2016) Song (2015) Lin (2016) Wang (2012) Zhao (2013) Zhang ( reductase (DHFR), and glycinamide ribonucleotide formyltransferase (GARFT) [26][27][28]. Several prospective studies have demonstrated that Pemetrexed has good tumor response rate and less chemotherapy related toxicity as the second-line chemotherapy for advanced NSCLC. Gefitinib is the first selective inhibitor of epidermal growth factor receptor's (EGFR) tyrosine kinase domain which was approved by the FDA for the treatment of patients with locally advanced or metastatic NSCLC after failure of both platinum-based and Docetaxel chemotherapies [29,30]. A large-scale, double-blinded prospective randomized clinical trial (IRESSA Pan-Asia Study, IPASS) [31] comparing Gefitinib versus carboplatin/paclitaxel in treatment of advanced NSCLC was published in The New England Journal of Medicine in 2009, and showed the PFS was significantly longer for Gefitinib than chemotherapy in patients with EGFR mutation positive tumors. However, the PFS was significantly longer for chemotherapy than Gefitinib in patients without EGFR mutation.
However, the clinical treatment response and drug related toxicity of Pemetrexed versus Gefitinib as secondline treatment for advanced NSCLC was not clear although several studies had investigated this topic. In our present work, we performed a meta-analysis by pooling the data from 14 openly published studies. The combined results show that Pemetrexed is not inferior as second-line treatment for advanced NSCLC compared to Gefitinib for tumor response. However, Pemetrexed has higher incidence of neutropenia but lower risk of developing skin rash and diarrhea. In the clinical practices, the cancer tissue could not be obtained in all the advance NSCLC patients. Therefore, the EGFR mutation status was not clear for a large part of the patients. For the patients with unknown EGFR status, Pemetrexed was a rational selection as a second-line chemotherapy.
Our work demonstrated that Pemetrexed was not inferior as second-line treatment for advanced NSCLC compared to Gefitinib in tumor response. However, Pemetrexed had higher incidence of neutropenia but lower risk of developing skin rash and diarrhea. Our work also has limitations: (i) Only 14 clincal trials were included in the meta-analysis, the sample size is relative small. (ii) Significant statistical heterogeneity for the ORR and DCR existed, which may decrease the statistical power. (iii) Only studies published in English or Chinsese were inlcuded. Therefor, the conclusion of this work needs further verification by well designed RCTs or high quality individual meta-analysis.

Conflict of interest: Authors state no conflict of interest.
Overall (I-squared = 0.0%, p = 0.914) Dai