Effectiveness, Uses and Safety of Granulocyte Colony Stimulating Factor biosimilars: a protocol for systematic review and meta-analysis

Background Granulocyte Colony-stimulating factors (G-CSF) biosimilars are recombinant biologics that are similar to a reference product, neupogen, a 175 amino acid recombinant human G-CSF. Characteristically, biosimilars are produced from living cells as high molecular weight, heterogenous compounds that are highly immunogenic. They are primarily used to treat febrile and severe neutropenia in oncology patients as well as mobilize peripheral stem cells in transplant donors. However, as biosimilars are produced by biological processes rather than chemical synthesis, their comparable effectiveness and safety are very paramount to clinical uses. We aimed to produce a protocol for consistent and accurate systematic review and meta-analysis of G-CSF biosimilars. Methods We developed a search strategy using MeSH terms, key words and entry terms to search 9 databases: PubMed, AJOL, Embase, Google Scholar, Scopus, Cochrane Library, CINAHL, Web of Science and ResearchGate. Only randomized controlled trials retreivable in the English language will be included in this study. The primary measurable outcomes in this study are uses, effectiveness and safety of G-CSF biosimilars. Identied primary studies will be screened, deduplicated and selected based on study design, inclusion/exclusion criteria and outcome measures using DistillerSR software. Studies will be assessed for methodological, clinical and statistical heterogeneity. Extractable data items for effectiveness measure are: i) proportion of patients with 50% increase in absolute neutrophil count within 3–7 days; ii) resolution of fever within 3–7 days, iii) resolution of intra-oral mucosa ulcers within 7–10 days and iv) resoluton of diculty in swallowing within 7–10 days. Measures of safety are i) proportion of patients with immunologic reactions, ii) any other documented adverse events. Quality scores and risk of bias for individual studies will be reported. Funnel Plots will be used for assessing publication bias in selected studies. Effect size, variance, SE and % CI and heterogeneity tests will be reported on forest plots using the CMA software version 3. Subgroup analysis and meta-regression will also be included using secondary outcomes as moderators and explanatory variables. The systematic review and meta-analysis will be reported according to PRISMA 2015 Statement. Discussion Ethical approval will not be required since this study will be based on published data. G-CSF biosimilars will be compared with the reference product, neupogen (lgrastin). The uses, effectiveness and safety of the biosimilars will be discussed. The study will also examine short and long acting biosimilars and compare their overall effectiveness. The strength of evidence from this study will be assessed using the NIH Quality assessment for systematic review and meta-analysis.


Background
The uses and effectiveness of Granulocyte Colony-stimulating factors (G-CSF) in the management of haematological malignancies, solid tumours and mobilization of stem cells from peripheral blood have long been established [1][2][3]. Both the American Society of Clinical Oncology (ASCO) and the European Organization for Research and Treatment of Cancer (EORTC) guidelines recommend the use of G-CSF for the treatment and prevention of febrile neutropenia (FN) in haematologic malignancy and other oncology cases in patients undergoing chemotherapy [4,5]. G-CSF is the primary regulator of proliferation and differentiation, maturation, survival, and functions of granulocytes to exert the biological defense mechanism [6,7]. The discovery of lgastrim [8] and the approval of its biosimilars [9,10] have revolutionized cancer treatment leading to increased survival of patients undergoing chemotherapy.
G-CSFs are a type of haemopoietic growth factors which are synthesized mainly in the bone marrow to aid stimulation, proliferation, differentiation, and survival of neutrophil precursors in the bone marrow and to promote their maturation process. Its use gain popularity among cancer patients who had chemotherapy-or radiotherapy-induced neutropenia where it reduced the degree and duration of neutropenia [7,10]. It has been used for other causes of severe neutropenia. It is also used in patients undergoing stem cell transplantation to mobilize peripheral stem cells. G-CSF exerts its effect on its target by binding to and forming complexes with the haematopoietin receptor family. Once bound to the receptors, they form G-CSF-Receptor ligand complex which causes dimerization of the receptor and signal transduction through proteins including Jak, Lyn, STAT and Erk1/2 [11].
In 2016, G-CSF and other biologics contributed signi cantly to global expenditure on medicare, causing increase in healthcare spending [12,13]. Biosimilars are biologics with high similarity but not identical to a reference drug. They are believed to offer some reduction in global cost of drugs with a potential of saving up to 22% of the cost on the biologics over a 10-year period in the United States [14,15]. Apart from reduction in cost, biosimilars have become readily available and accessible to many patients [13]. G-CSF biosimilars are already in use and they show similar effects with G-CSF [16]. Generally, biosimilars are usually subjected to structural, biochemical and clinical comparative analysis to the reference products by regulatory bodies [17][18][19][20].
G-CSF biosimilars are recombinant biologics that are similar to a reference product, neupogen, a 175 amino acid recombinant human G-CSF [21]. They are classi ed into a) shorting acting such as the lenograstim, lgrastim which includes neupogen, zarzio, nivestim and acco l; and b) the long-acting G-CSFs such as the pegylated lgrastim and lipeg lgrastim. Presently, European Medicines Agency has approved about twenty-one G-CSF biosimilars for clinical use [22] while the FDA approved peg lgrastim (Neulasta) in 2002 as a long-acting form of G-CSF [23]. FDA also approved three biosimilars: peg lgrastim-jmdb (Fulphila), peg lgrastim-cbqv (Udenyca; Coherus BioSciences) [24] and peg lgrastim-apgf (Nyvepria) [25]. Characteristically, biosimilars are produced from living cells as high molecular weight, heterogeneous compounds that are highly immunogenic [26]. Although they are di cult to fully characterize, their e cacy, safety and uses are similar. Their supply is highly dependent on stringent manufacturing processes that are very exclusive to each product. Overall, biosimilars have reduced cost of medicare for cancer patients. Fundamentally, biosimilars are different from generics, a term used mainly for off-patent chemically synthesized drugs that are similar to a parent drug. Generics have a well-de ned chemical structure; they are small molecules with identical copies in the manufacturing process [27]. They are mostly non-immunogenic and very stable to external variations. With increasing approvals of G-CSF biosimilar medicines, by both FDA and EMA, some new and promising biosimilars may still not receive approval on reasons of lack of optimal effectiveness and safety.
Generally, G-CSF biosimilars are indicated for neutropenia and/or peripheral stem cell mobilization. We aim to develop a protocol to enable a reliable, robust and accurate systematic review and meta-analysis of the uses, effectiveness and safety of G-CSF biosimilars.

OBJECTIVE
The overall objective of this study is to evaluate the effectiveness, uses and safety of G-CSF biosimilars in oncology patients undergoing chemotherapy, nononcology patients with severe neutropenia and peripheral mobilization of stem cells.
The study will determine the effects and impacts of long-acting G-CSF compared to that of short acting forms in reducing neutropenic fever in oncology patients undergoing chemotherapy.
Study objectives: 1. To determine the overall effectiveness of G-CSF biosimilars. i. Interventional studies: only RCTs and quasi-clinical trials in humans will be included.
ii. Studies are retrievable in the English language.
iii. Studies report the primary outcome with or without secondary outcomes.
c. Exclusion Criteria: The speci c criteria for excluding studies are: i. Observational studies, reviews, comments and letters to editors.
ii. Studies that do not report primary outcome.
v. Studies not retrievable in the English Language.

d. PICOs
Population: Individuals including oncology patients who are neutropenic and, or prophylactic use for peripheral stem cell mobilization in transplant donors.
Outcomes: The primary outcomes are effectiveness, uses and safety of G-CSF biosimilars. The secondary outcomes include overall improvement in cancer patients' quality of life, rationale for short vs long acting biosimilars and types of biosimilars approved and available in the market.

Information Sources
Our search uses sensitive topic-based strategies designed for each database. Nine databases will be searched: PubMed, CINAHL, Embase, Web of Science, Google Scholar, AJOL, Cochrane library, ResearchGate and Scopus will be included. Only randomized clinical trials and quasi-clinical trials, retrievable in the English langauges will be included.

Search strategy
The search strategy will include MeSH terms, text words, and entry terms. The search strategies used in databases are shown in Table 1.

Data Extraction and Management a. Data Extraction
Data will be managed in three main softwares: DistillerSR, CMA version 3 and Microsoft Excel.
Identi ed studies will be screened independently in pairs and blindly using the DistillerSR software at 6 different levels: i. Level 1 would involve screening of identi ed studies for the study design. Only interventional studies would be accepted ii. Level 2 will involve screening of identi ed studies in the titles and abstracts using the prede ned study characteristics iii. Level 3 will involve further screening of the contents of articles by reading the full text articles using the search strategy. iv. Level 4 will involve snowballing of literature on references from included studies.
v. Level 5: Studies will be screened for outcome measures of effectiveness, uses and safety of G-CSF biosimilars.
vi. Level 6: studies will be assessed for risk of bias using NIH Quality Assessment for Interventional studies and Cochrane Risk of Bias for Randomized Clinical trials.
Con icts during screening will be resolved by a third independent reviewer who serves as a tie breaker.

a. Selection Process:
Screened studies will be selected based on study characteristics: study design, inclusion/exclusion criteria and agreement between two independent and blinded reviewers. Authors of included studies with missing data will be contacted via email and telephone. After selection, studies will be deduplicated. Data items will be extracted from selected studies into prede ned forms in the DistillerSR. Data items will be exported into prede ned format in Microsoft Excel, to be imported into the CMA software for quantitative analysis.
The effect size for the primary outcome is prevalence. The effect sizes for secondary outcomes are some categorical and others quantitative.

Risk of bias
The risk of bias in the selected studies will be assessed for each study using the NIH Quality assessment tool for interventional studies. The NIH Quality assessment tool has 14 questions, with scores about 8 indicating good quality study. This will be cross-checked with the Cochrane tool of risk of bias assessment for randomized clinical trials. Publication bias in the selection of studies will be visually assessed using the funnel plot (trim and ll method) and test for asymmetry. Other statistical tests such as Egger's regression intercept, Begg and Mazumdar's rank correlation and Orwin's fail-safe N will be used where appropriate. Studies with extreme bias will be subjected to sensitivity testing using the include/exclude function in the CMA Software.

Assessment of Meta-bias
Meta-bias will be assessed as follows: i. Method of testing/reporting of effectiveness and safety measures for biosimilars. This will be done at the outcome level.
ii. Index of reporting measurable outcomes in included studies: Studies that were reported in different indices but similar in outcome and design will be converted to the primary effect size, prevalence, based on individual case evaluation.
iii. Heterogeneity will be assessed at the study level using the Q statistics, and its p-value, I², ² (Tau squared). As a rule of thumb, I² values of less than 40% will be considered low heterogeneity while values > 40 but < 75 % will be considered moderate and values > 75% are high. This is done at the study level.

Data synthesis
Criteria for Data Synthesis Extracted data items, that reported primary outcomes with or without secondary outcomes will be used for both narrative synthesis and quantitative analysis.
The following criteria will be applied for analysis: a. Studies that passed the methodological quality assessment using the NIH quality assessment tool, crosschecked with the Cochrane Risk of Bias tool will be included. The results will be presented in tabular format, indicating all the extractable data items and each study quality score.
b. All studies with primary outcomes will be used for narrative synthesis.
c. All studies with primary outcomes and secondary outcomes that pass heterogeneity tests and have good quality scores will be used for quantitative synthesis.
d. Further Analysis: Subgroup analysis will be performed using variables such as type of biosimilars and rationale for choice of short vs long acting biosimilars as moderators. Meta-regression will be performed on quantitative variables such patient's quality of life, proportion of patients with 50% increase in absolute neutrophil count with 3-7 days (dependent variable) and other quantitative measures of effectiveness.
e. Where heterogeneity exists, sensitivity testing using include/exclude functions in the CMA software will be performed.
f. Both xed and random computational models will be used to compare pooled/ summary prevalence, standard error, variance and 95% CI.
Presentation and Reporting of Results: The pooled prevalence, standard error, variance, 95% CI of effectiveness, uses and safety of G-CSF biosimilars will be reported in forest plots. The relative weights assigned to studies will be reported in forest plots as well. Both NIH quality score and the Cochraine risk of bias will be presented in tables along with extracted measurable outcomes. Tests of heterogeneity will be included in the forest plots. Subgroup analysis will also be reported in forest plots. Metaregression and publication bias analysis will be presented in plots while results of sensitivity testing will be reported in a table.
The study is funded by the Molecular Pathology Institute, Enugu Nigeria.

Support
The Molecular Pathology Institute provided the DistillerSR via the Institute subscription.