1 Introduction
Pancreatic cancer (PC), which is derived from the glandular tissue of the pancreas, is a devastating malignant tumor and characterize with high mortality and poor prognosis.[1] And the incidence of PC is increasingly rising. PC is the fourth major cause of cancer-related death in all population worldwide; it causes about 338,000 new cases each year.[2] The 5-year survival rate of PC patients with R0 pancreatic surgery is 6%, and the median overall survival time is 4 to 6 months in patients with metastatic disease.[3] Although surgical resection may offer a good prognosis and prolong survival, approximately 80% patients with PC are always diagnosed as unresectable tumor (locally advanced and/or metastatic), because the symptoms of PC generally occur late, and it will lead to the extremely poor prognosis for advanced PC.
Chemotherapy is considered as a major treatment for unresectable PC, and according to the National Comprehensive Cancer Network's (NCCN) recommendation, gemcitabine-based chemotherapy is a standard backbone for advanced/metastatic PC.[4,5] However, recent studies showed that better therapeutic options are available for advanced/metastatic PC, such as target therapies.[6,7] Both erlotinib (Tarceva), everolimus (Afinitor) and sunitinib (Sutent) were global approved for management and treatment of advanced PC and prolonged progression-free survival (PFS) for PC patients. Sunitinib malate is an oral small-molecule tyrosine kinase inhibitor.[8–10] Erlotinib, an orally bioavailable inhibitor of EGFR, has been clinically approved for unresectable pancreatic cancer in combination with gemcitabine.[11] Everolimus as the mammalian target of rapamycin inhibitor is demonstrated that can significantly prolong PFS in patients with advanced PC.[12]
Wang's meta-analysis focused on gemcitabine plus erlotinib for locally advanced/metastatic PC. Vinik's systematic review and meta-analysis focused on sunitinib for the quality of life of advanced PC patients.[13] And everolimus may prolong PFS and down-regulates excess production of 2 gastrointestinal hormones in patients with PC.[14] However, there was no meta-analysis to compare the efficacy of different targeted agents for PC patients. Thus, we cannot determine which targeted agent is superior to other treatment using randomized controlled trial or pairwise comparison meta-analysis.
In the network meta-analysis, the available information from pairwise comparisons of treatment A and treatment B is combined with indirect comparisons C either a third intervention or a control condition to estimate the relative effectiveness among all interventions and rank ordering of the interventions even if head-to-head comparisons are lacking.[15]
Present systematic review and network meta-analysis will evaluate the relative efficacy of different targeted agents combined with chemotherapy for advanced/metastatic PC in the improvement of overall survival, progression-free survival, and adverse events using Bayesian network meta-analysis.
2 Methods
This protocol will be reported according to preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P).[16] The study protocol has been registered on the international prospective register of systematic review (PROSPERO) (CRD42017076728).
2.1 Eligibility criteria
Studies will be included in this systematic review and network meta-analysis if meet the following eligibility criteria: randomized controlled trials (RCTs); patients from all countries and the age is more than 18; patients that are diagnosed with advanced/metastatic PC; studies focused on the differences between different targeted agents and/ or chemotherapy; the outcome is progression-free survival, overall survival, and adverse events; there will be no limitations on year of publication, publication status, and language of publication.
2.2 Literature search and study selection
To identify relevant studies, the following 6 electronic databases will be searched: PubMed, EMBASE, the Cochrane Central Register of Controlled Trials (CENTRAL), Web of science, CNKI (Chinese National Knowledge Infrastructure), and CBM (Chinese Biological Medical Database). Also we will manually search the bibliographies of included articles and relevant systematic reviews and meta-analysis to identify other additional studies.
The following search terms will be used: pancreatic neoplasm, pancreatic cancer, cetuximab, erlotinib, tarceva, everolimus, afinitor, sunitinib, sutent, bevacizumab, trastuzumab, trametinib, ganitumab, and ruxolitinib.
Two authors will independently screen the title and abstract of retrieved studies. Moreover, the potentially eligible studies will be assessed by retrieving the full texts. In addition, a third reviewer will be requested in case of disagreement. We will use a flow diagram to illustrate the study selection process according to PRISMA guidelines.[17]
2.3 Data collection
2.3.1 Data management
We will first perform a pilot test between 2 reviewers to ensure high inter-rater reliability. Then the management of literature search records will be conducted in ENDNOTE X7.
2.3.2 Data extraction
A standard data extraction form will be created using Microsoft Excel 2013 (Microsoft Corp, Redmond, WA, www.microsoft.com) to collect data of interest, which including general characteristics of included trials (e.g., name of first author, year of publication, whether single-center or multicenter, country of study, recruitment time frame, follow-up length, total sample size, inclusion and exclusion criteria), details of participants (e.g., gender, age, tumor stage, tumor size, lymph node status), details of interventions (e.g., regimens of interventions, dosage), and outcomes.
2.3.3 Quality assessment
The quality of a body of evidence will be assessed by paired reviewers with the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) approach.[18] Any conflicts will be resolved by consulting an independent adjudicator. Direct evidence from RCTs will start at high quality and can be rated down based on risk of bias, indirectness, inconsistency, and publication bias. The rating of indirect estimates will start at the lowest rating of the 2 direct estimates that contribute as the first-order loops to the indirect estimate but will be rated down further for intransitivity. If direct and indirect estimates contribute similar power to the network estimate, then we will use the higher rating as the rating of network meta-analysis. The network meta-analysis will be further rated down if they are inconsistency and imprecision.
2.3.4 Risk of bias of included studies
The risk of bias of included studies will be estimated using the Cochrane Handbook version 5.1.0[19] tool, which includes 7 specific domains: sequence generation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias and detection bias), incomplete outcome data (attrition bias), selective outcome reporting (reporting bias), and other bias. Based on criteria for judging the risk of bias,[19] we will classify methodological quality as low risk of bias “+,” high risk of bias “−,” or unclear risk of bias “?” Two independent reviewers will complete the assessment of risk of bias. The conflicts will be resolved by a third reviewer.
2.3.5 Data synthesis
We will use Microsoft Excel 2013 to design a form summarize data of all the included studies and showing their major characteristics and some important information related to this systematic review and meta-analysis. We will use STATA version 12.0 software to combine data and conduct a pairwise meta-analysis. I2 statistic will be used to conduct heterogeneity assessment. If 0≤I2≤25%, we consider statistical heterogeneity as small; as medium if 25% < I2≤50%; as large if I2> 50%.[20] Random-effect model will be used if the heterogeneity exists, otherwise, fixed-effect model analysis will be performed. We will use pooled odds ratios (ORs) with 95% confidence interval (95%CI) to show dichotomous outcomes, and use standard mean difference (SMD) or weight mean difference (WMD) with 95% CI for continuous outcomes.
We will use R-3.4.1 software and package gemtc version 0.9-2 to conduct a Bayesian network meta-analysis.[21] We will use node splitting method to evaluate inconsistency between direct and indirect comparisons if a loop connecting 3 arms exists. The treatment ranking will be presented based on the point estimates and standard errors of the network assess.
2.3.6 Subgroup and sensitivity analyses
Considered of possible significant heterogeneity or inconsistency, we will use subgroup analysis to find the possible sources. We also will estimate the sensitivity of results according to the results of risk of bias.
2.3.7 Publication bias
We will use STATA version 12.0 software (Stata Corporation, College Station, TX) to draw a comparison-adjusted funnel plot to identify whether there will be a small sample effect among the networks.
3 Author contributions
DBS, GL, CN, and GTK planned and designed the research; GL, PB, and MJC tested the feasibility of the study; DBS and PB wrote the manuscript; all authors approved the final version of the manuscript.
Conceptualization: B. Di, L. Ge, T. Guo.
Data curation: B. Pan, J. Ma.
Investigation: B. Di, B. Pan, J. Ma, Y. Wu.
Methodology: B. Di, L. Ge.
Supervision: T. Guo.
Writing – original draft: B. Di, B. Pan.
Writing – review & editing: B. Di, J. Ma, L. Ge, T. Guo, Y. Wu.
References
[1]. Ferla J, Soerjomataram I, Ervik M, et al. GLOBOCAN 2012 v10, Cancer Incidence and Mortality Worldwide: IARC Cancer Base No. 11. Int J Cancer 2014;36:E359–86.
[2]. Bond-Smith G, Banga N, Hammond TM, et al. Pancreatic adenocarcinoma. BMJ 2012;344:e2476.
[3]. Siegel R, Ma J, Zou Z, et al. Cancer statistics. 2014. CA Cancer J Clin 2014;64:9–29.
[4]. Burris HA, Moore MJ, Andersen J, et al. Improvements in survival and clinical benefit with gemcitabine as first-line therapy for patients with advanced pancreas cancer: a randomized trial. J Clin Oncol 1997;15:2403–13.
[5]. Arslan C, Yalcin S. Current and future systemic treatment options in metastatic pancreatic cancer. J Gastrointest Oncol 2014;5:280–95.
[6]. Hao J, Yang X, Ding XL, et al. Paeoniflorin potentiates the inhibitory effects of erlotinib in pancreatic cancer cell lines by reducing ErbB3 phosphorylation. Sci Rep 2016;6:32809.
[7]. Lee JM, Lee HS, Hyun JJ, et al. Prognostic value of inflammation-based markers in patients with pancreatic cancer administered gemcitabine and erlotinib. World J Gastrointest Oncol 2016;15:555–62.
[8]. Abrams TJ, Lee LB, Murray LJ, et al. SU11248 inhibits KIT and platelet derived growth factor receptor beta in preclinical models of human small cell lung cancer. Mol Cancer Ther 2003;2:471–8.
[9]. Kodera Y, Katanasaka Y, Kitamura Y, et al. Sunitinib inhibits lymphatic endothelial cell functions and lymph node metastasis in a breast cancer model through inhibition of vascular endothelial growth factor receptor 3. Breast Cancer Res 2011;13:R66.
[10]. Mendel DB, Laird AD, Xin X, et al. In vivo antitumor activity of SU11248, a novel tyrosine kinase inhibitor targeting vascular endothelial growth factor and platelet-derived growth factor receptors: determination of a pharmacokinetic/pharmacodynamic relationship. Clin Cancer Res 2003;9:327–37.
[11]. Moore MJ, et al. Erlotinib plus gemcitabine compared with gemcitabine alone in patients with advanced pancreatic cancer: a phase III trial of the National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 2007;25:1960–6.
[12]. Martins D, Spada F, Lambrescu I, et al. Predictive markers of response to everolimus and sunitinib in neuroendocrine tumors. Target Oncol 2017;5:611–22.
[13]. Vinik A, Bottomley A, Korytowsky B, et al. Patient-reported outcomes and quality of life with sunitinib versus placebo for pancreatic neuroendocrine tumors: results from an international phase III trial. Target Oncol 2016;11:815–24.
[14]. Pavel ME, Chen D, He W, et al. Everolimus effect on gastrin and glucagon in pancreatic neuroendocrine tumors. Pancreas 2017;46:751–7.
[15]. Bafeta A, Trinquart L, Seror R, et al. Reporting of results from network meta-analyses: methodological systematic review. BMJ 2014;348:g1741.
[16]. Shamseer L, Moher D, Clarke M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P): 2015: elaboration and explanation. BMJ 2015;349:g7647.
[17]. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg 2010;8:336–41.
[18]. Puhan MA, Schunemann HJ, Murad MH, et al. A GRADE Working Group approach for rating the quality of treatment effect estimates from network meta-analysis. BMJ 2015;350:h3326.
[19]. Higgins JPT, Altman DG, Sterne JAC (editors). Chapter 8: Assessing risk of bias in included studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from
www.handbook.cochrane.org.
[20]. Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002;21:1539–58.
[21]. Rücker G, Schwarzer G, Krahn U, König J. netmeta: Network meta-analysis using Frequentist methods. Available at:
http://cran.r-project.org/web/packages/netmetapdf. Accessed December 6, 2016.
