The Efficacy of Brucea javanica Oil Emulsion Injection as Adjunctive Therapy for Advanced Non-Small-Cell Lung Cancer: A Meta-Analysis

Purpose. To evaluate the efficacy of Brucea javanica oil emulsion injection (BJOEI) in patients with advanced non-small-cell lung cancer (NSCLC) during chemotherapy. Method. Electronic database of EMBASE and PubMed and the conference proceeding of ASCO, CNKI, CBMdisc, VIP, and Wanfang database were searched to select RCTs comparing BJOEI plus chemotherapy with chemotherapy alone in the treatment of advanced NSCLC, until June 1, 2016. Two reviewers independently performed the analysis according to the inclusion and exclusion criteria. Review Manager 5.3 and STATA 12.0 were employed for data analysis. Result. Twenty-one studies including 2234 cases were included. The pooled result indicated that there were significant differences in ORR (RR = 1.25; 95% CI: 1.14–1.36; P < 0.00001), improvement of QOL (RR = 1.87; 95% CI: 1.63–2.15; P < 0.00001), nausea and vomiting (RR = 0.67; 95% CI: 0.46–0.98; P = 0.04), leukopenia (RR = 0.63; 95% CI: 0.52–0.75; P < 0.00001), but there was no difference in thrombocytopenia (RR = 0.78; 95% CI: 0.49–1.23; P = 0.29). Begg's funnel plot and Egger's test indicated that no publication bias was found. The sensitivity analysis suggested the stability of the pooled result. Conclusion. The addition of BJOEI can enhance efficacy, improve QOL, and decrease incidence of nausea and vomiting and leukopenia for advanced NSCLC patients. However, higher quality RCTs are needed to further confirm this finding.


Introduction
As a prevalent and highly malignant carcinoma, lung cancer is the leading cause and the second leading cause of cancer death among men and women worldwide, respectively [1]. While the incidence of lung cancer has declined in some regions like North America and Europe, it is still high in China with the incidence rate of 733.3 per 100,000 in 2015 [2]. Moreover, with an annual growth rate of 26.9%, the expected number of cases of lung cancer will reach 100 million in 2025, indicating that China might have the largest population of lung carcinoma patients around the world [3].
Lung cancer can be divided into two types: one is smallcell lung cancer and the other is non-small-cell lung cancer. It is estimated that the latter, of which the most common types are squamous cell carcinoma, large-cell carcinoma, and adenocarcinoma, accounts for 80%∼85% of all global lung cancer cases [4]. The treatments of non-small-cell lung cancer (NSCLC) give first place to surgery, but most of the patients cannot receive the appropriate resection as they have reached an advanced stage when being diagnosed [5]. Standard firstline treatment for advanced NSCLC involves a combination of two drugs, including a platinum compound and a nonplatinum compound, such as paclitaxel and docetaxel, having indeed achieved favorable outcome [6]. However, the substantial toxicity incurred by chemotherapy, including gastroenteric reaction, hematotoxicity, nausea, and vomiting, should not be overlook, which is negatively correlated with quality of life (QOL) for NSCLC patients [7], especially for the advanced ones. Thus, how to reduce the burden of toxicity and achieve higher quality of life is the top priority on the clinical research agenda [8].
Traditional Chinese Medicines (TCMs) have become increasingly popular in the treatment of cancer in China [9]. Brucea javanica oil emulsion injection (BJOEI) is one of TCMs products, which takes Brucea Jen petroleum ether 2 Evidence-Based Complementary and Alternative Medicine extracts as raw material and purified soybean lecithin as emulsifier [10], and is employed as adjunctive therapy in the treatment of lung carcinoma, brain metastasis of lung carcinoma, and gastrointestinal tumorigenesis. A great number of published studies have proved that BJOEI can perform a synergetic antitumor effect by improving tumor response, boosting Karnofsky Performance Score (KPS), reducing the incidence of adverse events and stimulating the immunity during chemotherapy or radiotherapy [11]. The pooled result of a meta-analysis showed that the addition of BJOEI to chemotherapy produced favorable outcomes for patients with advanced gastric cancer, including improvement of objective response and QOL and reduction of side effects such as neutropenia, thrombocytopenia, nausea, and vomiting [12].
BJOEI has been applied in clinical practice for advanced NSCLC patients since long time ago, but no relevant metaanalysis was conducted. Thus, we perform this meta-analysis to investigate the clinical efficacy of BJOEI plus chemotherapy in the treatment of advanced NSCLC.

Literature Source and Search Strategy.
We searched and extracted eligible studies about BJOEI treatment of NSCLC from databases of PubMed, EMBASE, the conference proceeding of American Society of Clinical Oncology (ASCO), Chinese Biological Medical disc (CBMdisc), Chinese National Knowledge Infrastructure (CNKI), Chinese Scientific Journals Full-text Database (VIP), and Wanfang database. The key words applied in the search were as followed: "lung cancer", "non-small-cell lung cancer", "NSCLC", "Brucea javanica oil emulsion", "BJOEI", "Yadanzi", and "chemotherapy". The retrieved studies were regarded as potential source and reviewed manually. Moreover, although the published year of these literatures were unlimited, only English and Chinese literatures were accepted.

Inclusion and Exclusion
Criteria. The following criteria were used for the literature inclusion. (1) The study design was confined to randomized controlled trials (RCTs) comparing platinum-contained chemotherapy alone with platinumcontained chemotherapy plus BJOEI for the NSCLC. (2) Study subjects who (a) were patients with stage III or IV NSCLC diagnosed pathologically and (or) cytologically; (b) had KPS ≥ 60 and (or) time of survival ≥ 3 months; (c) had outcomes of objective response rate (ORR) determined by World Health Organization (WHO) criteria or Response Evaluation Criteria in Solid Tumors (RECST), improvement of QOL evaluated by KPS, and adverse reactions assessed by WHO Recommendations for Grading of Acute and Subacute Toxicity; (d) had no chemotherapy contraindication before treatment and no significant abnormalities in liver, kidney, and heart function. The major exclusion criteria were as follows: (a) non-RCTs studies; (b) animal experiments, review, and other irrelevant studies; (c) no detailed data about ORR, improvement of QOL, and adverse events or no indicators for them; (d) single-arm study.

Endpoint Indicator.
The outcomes included clinical efficacy, quality of life, and adverse effects. According to WHO criteria and RECST, the tumor response included complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD). The ORR was defined as CR + PR. Toxicity was graded from 0 to IV in severity on the basis of the WHO Recommendations. This meta-analysis only investigated the incidence of Grade II or above nausea and vomiting, leukopenia, and thrombocytopenia.

Data Extraction and Quality Assessment.
Two reviewers independently extracted the information of the included study including name of author(s), publication year, number of patients in BJOEI group and control group, age, sex, chemotherapy regimen, stage of cancer, initial KPS, method of outcome ascertainment, study outcome, and detail of BJOEI treatment. Disagreement and problems were resolved by discussing or consulting with another reviewer according to the Cochrane handbook. The general methodological quality of each included trials was assessed by six items according to the Cochrane Collaboration's Risk of Bias (ROB) criteria. The items are about randomization, allocation concealment, double blinding, integrity of outcome data, selective reporting, and other bias. Every item is given a possible score of 0 for low, 1 for medium, and 2 for high ROB, all yielding a total score ranging from 0 to 12 for each study. Low ROB is appointed to trials with total score from 0 to 4, medium ROB with total score from 5 to 8, and high ROB with total score from 9 to 12.

Statistical
Analysis. All of the data was calculated by STATA 12.0 software package and Review Manager 5.3 software. The risk ratio (RR) with 95% CI was applied to analyze the dichotomous data. 2 and 2 tests were used to assess statistical heterogeneity among included studies. If there was no heterogeneity across the trials, the pooled result was obtained by the fixed effect model; otherwise, the random effect model was used. Sensitivity analysis was conducted to estimate the stability of pooled result. Further, we employed Begg's funnel plot and Egger's test to test the publication bias.

Search Result.
We conducted the systematic research on June 1 , and 251 potentially relevant references were yielded from online database. A total of 90 articles were excluded for duplication and 161 studies were entered next step. After further screening and eligibility assessment, 140 trials were excluded. Finally, 21 trials were selected as appropriate for inclusion in this meta-analysis. The flow chart showing the selection process was presented in Figure 1.

Objective Response Rate.
All included studies reported ORR in each arm. The heterogeneity analysis showed that no significant heterogeneity was found ( 2 = 0.00%; = 0.98), and we applied the fixed effect model in the pooled analysis. As shown in Figure 2, the pooled result indicated a better RR in BJOEI treatment group than in control group (RR = 1.25; 95% CI: 1.14-1.36; < 0.00001). The results in subgroup analysis of GP regimen (RR = 1.35; 95% CI: 1.14-1.59; = 0.0004) and DP regimen (RR = 1.25; 95% CI: 1.08-1.45; = 0.003) also demonstrated the favorable outcome. However, there were no significant differences with regard to ORR between BJOEI group and control group in subgroup analysis of both TP regimen and NP regimen. The integrated RR for ORR in TP regimen subgroup was 1.17 (95% CI: 0.89-1.54; = 0.26) and the pooled RR for NP regimen group was 1.14 (95% CI: 0.94-1.37; = 0.18). Briefly, this metaanalysis indicated BJOEI plus chemotherapy improved tumor response.

Improvement of QOL.
Twelve studies reported the improvement of KPS. No significant heterogeneity was found among these studies ( 2 = 0.00%; = 0.93); thus we employed the fixed effect model in this meta-analysis. The pooled result demonstrated that BJOEI combined with chemotherapy could significantly improve the QOL (RR = 1.87; 95% CI: 1.63-2.15; < 0.00001), which was illustrated in Figure 3.

Grade II or above Nausea and Vomiting.
Six studies provided information about nausea and vomiting of the BJOEI treatment group and chemotherapy alone group. The heterogeneity test demonstrated no significant heterogeneity among the studies ( 2 = 0.00%; = 0.67); thus we used the fixed effect model. As shown in Figure 4, the pooled results indicated that the BJOEI could decrease the risk of developing nausea and vomiting when patients received the chemotherapy combined with BJOEI (RR = 0.67; 95% CI: 0.46-0.98; = 0.04).
3.6. Grade II or above Leukopenia. There were eight studies that provided data on Grade II or above leukopenia. The heterogeneity among the included studies was not significant ( 2 = 26%; = 0.22). As 2 < 50%, they were considered 4 Evidence-Based Complementary and Alternative Medicine  to be homogeneous and a fixed effect model was employed for analysis ( Figure 5). The result showed that the BJOEI combined with the chemotherapy decreased the incidence of Grade II or above leukopenia (RR = 0.63; 95% CI: 0.52-0.75; < 0.00001).

Grade II or above Thrombocytopenia.
Four studies provided the data of Grade II or above thrombocytopenia in both arms with no statistical heterogeneity ( 2 = 0%; = 0.91).
Fixed effect model was employed in the meta-analysis. As illustrated in Figure 6, the pooled resulted showed that the BJOEI did not decrease the incidence of Grade II or above thrombocytopenia (RR = 0.78; 95% CI: 0.49-1.23; = 0.29).

Publication Bias.
As all the eligible studies reported the outcome of tumor response, we chose to test the potential publication bias among the trials on ORR by Begg's funnel plot and Egger's test. As displayed in Figure 7, the funnel plot was symmetric, suggesting that no evidence of publication bias was found. Moreover, Egger's test provided evidence for no significant publication bias with = 0.887.

Sensitivity Analysis.
We performed a sensitivity analysis by sequentially omitting one single study to estimate the summary effect. We conducted the sensitivity analysis on the parameter of ORR for all the studied provided data on it. As shown in Figure 8, the combined effect after exclusion was close to that before exclusion, suggesting that the pooled analysis result was stable.

Discussion
As a powerful statistical analysis, meta-analysis can yield integrated result from individual study which focuses on the same issue [34]. We performed this meta-analysis to assess the effect of BJOEI plus chemotherapy on tumor response, quality of life, and side effects for advanced NSCLC patients. Twenty-one studies providing data on BJOEI plus chemotherapy versus chemotherapy alone were identified and analyzed comprehensively.   combined with TP regimen and NP regimen, no significant differences ( = 0.26 and = 0.18, resp.) were found between the BJOEI group and control group. However, the pooled result ( < 0.00001) and the subgroup analysis of GP regimen ( = 0.0004) and DP regimen ( = 0.003) favored the BJOEI combined chemotherapy group. It seems that the sample sizes of TP regimen and NP regimen were too small to test validity. Besides, we evaluated the effect of chemotherapy regimen on quality of life but no significant difference was found, indicating that no matter which regimen the

BJOEI Events Total
Events Total  Brucea javanica oil (BJO) is the main ingredient in BJOEI. In vitro, BJO exhibited a potential ability to kill non-smallcell lung cancer cells [35]. The anticancer activity of BJO might be attributed to the following properties: inducing apoptosis [36], disturbing the cell cycle [36,37], disrupting the cellular energy metabolism, and depressing the expression of vascular endothelial growth factor. Though the precise mechanism of this anticancer drug is poorly understood, our meta-analysis, together with the previous comprehensive analysis [38], corroborated the efficiency of BJOEI in clinical practice. In Wang's meta-analysis, twenty-two studies fulfilled the inclusion criteria and the pooled results showed that the addition of BJOEI during chemotherapy for NSCLC significantly increased the objective response rate (RR = 1.31; 95% CI: 1.18-1.45; < 0.00001), improved the quality of life (RR = 1.78; 95% CI: 1.51-2.09; < 0.00001), enhanced the immune function, and decreased the incidence of gastroenteric reaction (OR = 0.59; 95% CI: 0.44-0.80; = 0.0007). Different criteria may lead to slightly different finding between these two meta-analyses, but the patients in BJOEI group did demonstrate superior objective tumor response, higher quality of life, and less side effects.
However, the ROB problem of included TCMs studies should be noticed. The quality assessment result indicated     TCMs trials [39][40][41]. Researchers should take responsibility for making registration of clinical trials [42], paying more attention to experiment design and methodological quality [43], and receiving education to write high quality report, so as to increase the credibility of TCMs studies.
Actually, limitations exist in our meta-analysis. First, we regarded EMBASE, PubMed, and the conference proceeding of ASCO as main sources of eligible studies, but all the included studies comparing BJOEI plus chemotherapy with chemotherapy alone for advanced NSCLC patients were searched in Chinese academic database. Second, double blinding and allocation concealment were not developed and implemented in all studies, incurring potential risks of selection bias and impairing the quality of this meta-analysis. Third, some of the studies which did not mention detailed characteristics of patients' age and gender distribution might also raise the risk of bias.

Conclusion
As one of the TCMs, BJOEI has been widely employed in China for many years. In recent year, the clinical practice indicated that the combination of BJOEI and chemotherapy not only improved the ORR and QOL, but also reduced the incidence of adverse events for the advanced NSCLC patients. Our meta-analysis did demonstrate and provide objective evidence to support the efficacy of BJOEI. Given that the quality of included studies is not high and there is a ROB problem in the meta-analysis, the anticancer effect of BJOEI should be further confirmed by higher-quality RCTs.