The Effect of Granulocyte Colony Stimulating Factor (G-CSF) on the Outcome of Assisted Reproduction Technology in Thin Endometrium Women: A Meta-Analysis

Background : To explore the effect of granulocyte colony-stimulating factor (G-CSF) on thin endometrium in women undergoing assisted reproduction. Methods : We performed a methodical search from their inception to December 2022 in various electronic databases containing PubMed, Cochrane Library, Embase, Web of Science, Scopus, in addition to a manual search. All journals concerning the effect of G-CSF on thin endometrium were found. Selected studies, collected data, and assessed risk of bias were conducted by two investigators under precise inclusion and exclusion criteria independently. We applied Revman 5.3 software to accomplish the Meta-analysis of qualified studies. Results : This research included 8 studies, including 6 randomized controlled trials and 2 non-randomized controlled studies, a total of 673 patients. Based on the meta-analysis, we noted that compared with the control group, G-CSF significant improved embryo implantation rate [risk ratio (RR) =1.91, 95% confidence interval (CI) (1.26, 2.91), p = 0.002] and clinical pregnancy rate [RR = 1.73, 95% CI (1.22, 2.45), p = 0.002]. Compared with the control group, the endometrial thickness in the G-CSF group had non-significant increase compared with that of the control group [mean difference (MD) = 0.81, 95% CI (–0.04, 1.67), p = 0.06], in randomized controlled trial (RCT) studies, subgroup analysis shows G-CSF group increased significantly [MD = 1.13, 95% CI (0.56, 1.67), p < 0.0001]. Conclusions : G-CSF may improve implantation rate and clinical pregnancy rate in assisted reproductive technology, and have the potential to increase the endometrial thickness among women with thin endometrium. However, the results of the included studies should be explained with caution due to their limited quantity and quality, and more studies of higher quality are demanded to prove these findings.


Introduction
Infertility is assessed to affect approximately 8-12% of husband and wife of childbearing age worldwide [1].Assisted reproductive technology (ART) has been widely used in the treatment of infertility over the past few decades.In 2014, about 1.93 million ART cycles were carried out in 76 countries around the world, frozen embryo transfer cycles increased by about 2/3 compared with 2010 [2].High quality embryos and endometrial receptivity are both critical for a successful pregnancy; however, endometrial receptivity may represent a key rate-limiting factor for pregnancy rates in ART.An important marker of endometrial receptivity is endometrial thickness; a thin endometrium is closely associated with a poor pregnancy outcome after embryo transfer [3][4][5].In ART, a thin endometrium is often defined as an endometrial thickness <7 mm; this condition affects 1.3-2.3% of women undergoing fertility treatment [5][6][7].Treating patients with a thin endometrium is a crucial challenge for patients undergoing ART.
Clinically, there are several different approaches that considered to be helpful to treat patients with a thin endometrium, including aspirin, human chorionic gonadotropin, sildenafil citrate, estrogen, antioxidant, vitamin E, alpha-tocopherol, pentoxifylline, nifedipine and so on [8].These treatments are selected based on the type of embryo transfer (ET) being planned.Nonetheless, embryo transfer cycles must be cancelled repeatedly due to some patients' endometrium is still unresponsive after using these therapies, even embryo transferred may fail in implantation [8].
Granulocyte colony stimulating factor (G-CSF) is a cytokine that belongs to a large family of cytokines [9].Previous studies have shown that G-CSF act vital function in endometrial growth and pregnancy.G-CSF may be involved in embryo implantation by regulating decidual macrophages and Th2 reaction [10]; G-CSF can induce the proliferation of endometrial epithelium and stromal cells by increasing the staining level of proliferating cell nuclear antigen, and promote the regeneration of thin endometrium [11]; G-CSF shown to stimulate the directed differentiation of stem cells at the site of injury and increase the thickness of the endometrium in rat models [11].Therefore, G-CSF may be an effective method to treat thin endometrium.Several clinical studies have investigated the use of G-CSF to treat women with a thin endometrium [12][13][14][15][16].However, inconsistent results have been reported with regards to the influence of intrauterine G-CSF infusion on pregnancy rate in clinical studies.Some studies have reported an improvement in pregnancy rate [12][13][14] while others did not [15,16].Therefore, in the present study, we conducted a meta-analysis to investigate the effect of G-GSF on thin endometrium during ART in the hope that this may provide useful guidance on how to prepare the endometrium for embryo transfer in patients with a thin endometrium.

Inclusion Criteria
(1) The style of the study was a randomized controlled trial (RCT) or non-RCT.(2) Study subjects were patients whose endometrial thickness was <7 mm during ART.(3) Intervention measures in the treatment group involved the intrauterine infusion or subcutaneous application of G-CSF while the intervention measures in the control group involved the intrauterine infusion of normal saline or a blank control.(4) The outcomes of the study included embryo implantation rate, clinical pregnancy rate, and/or endometrial thickness on the day of embryo transfer.(5) The studies were published and contained sufficient information on sample size and research results.

Exclusion Criteria
(1) The style of the study was an observational study, descriptive study, retrospective study, case report or review.
(2) The study subjects were animals.(3) The baseline of the treatment group and the control group was inconsistent.(4) Studies involved a defective design or inappropriate statistical methods and could not be corrected.(5) Duplicate studies.

Search Strategy
We searched PubMed, the Cochrane Library, Embase, Scopus and Web of Science from their inception to December 2022; we also conducted a manual search.3).

Data Extraction
Two reviewers carried out articles selected and data extracted independently.First, the reviewers assessed relevant studies by reading the title and abstract.Then, they read through the entire study to decide whether it met the inclusion criteria.Next, the two reviewers compared their results; if they disagreed, they would take advice from an expert in the field who acted as a third reviewer.Finally, the reviewers acquired data from each selected study based on a pre-prepared data collection table.The corresponding author was contacted to obtain data whenever there was an incomplete dataset.The reviewers collated a range of data: (1) basic data, including the first author's name and publication date; (2) study design; (3) factors used to assess the risk of bias; (4) essential patient characteristics, including the size of the treatment and control groups, nation, patient age and baseline endometrial thickness; (5) information relating to intervention strategies including the timing, method, dose and frequency of intrauterine G-CSF infusion; (6) outcome indicators, including clinical pregnancy rate, embryo implantation rate and endometrial thickness.

Bias Risk Assessment
The risk of bias for the RCTs included in this metaanalysis was evaluated by the Cochrane risk-of-bias tool [17].We also employed the methodological index for nonrandomized studies (MINORS) to evaluate the non-RCTs bias risk [18].

Statistical Analysis
The meta-analysis was carried out with Review Manager 5.3 (Cochrane Collaboration, Copenhagen, Denmark) [19].We used mean differences (MD) to describe continuous variables and relative risk (RR) to show categorical variables.We calculated point estimates and 95% confidence intervals (95% CIs) for each outcome measure.The meta-analysis was performed with random and fixed effects models according to the I 2 statistic.Funnel plots were not generated because only 8 studies were included in this metaanalysis.

Study Selection
Following careful selection and consideration of the inclusion and exclusion criteria, 8 studies were finally selected for the meta-analysis, including 6 RCTs [12][13][14][20][21][22] and 2 non-RCTs [15,16].A total of 673 patients were included in our research.Fig. 1 shows the number of selected studies and the reasons for exclusion at each step.

Basic Characteristics and Risk of Bias Assessment
The characteristics of the included studies are presented in Table 1 (Ref.[12][13][14][15][16][20][21][22]), including the number of patients, nation, mean age, baseline endometrial thickness, interventions, and outcome measures for each study These studies covered 6 countries (2 in Iran, 2 in China, and each in Portugal, Spain, Poland, India).A total of 8 eligible studies were included in this meta-analysis.5 studies reported the embryo implantation rate, 7 eligible studies reported the clinical pregnancy rate, 4 studies reported endometrial thickness.
Table 2 (Ref.[15,16]) shows the results derived from the assessment of methodological quality of non-RCTs.Analysis showed that the non-RCTs scored 22 (more than 12 points) and were considered to represent a low risk of bias.

Discussion
Successful assisted reproduction requires a receptive endometrium; however, studies have shown that the implantation rate is significantly lower in women with a thin endometrium [23].A thin endometrium has an unfavorable effect on reproductive outcome.Various therapeutic methods have been used for patients with a thin endometrium.Although there are many types of treatments, most existing methods only achieve a slight change in endometrial thickness and pregnancy outcome.Consequently, it is very challenging to treat women with a thin endometrium when considering ART [24].Previous studies have suggested that G-CSF may increase the embryo implantation rate and clinical pregnancy rate in women with a thin endometrium.G-CSF, as a type of cytokine, can affect bone marrow mesenchymal stem cells and decidual macrophages, thus influencing endometrial growth and development [25].Using a rat model of thin endometrium, a previous study showed that the intrauterine infusion of G-CSF could improve endometrial receptivity by regulating endometrial hyperplasia and angiogenesis [26].Zhao et al. [27] reported that G-CSF could mobilize stem cells to the site of injury and repair tissue in animal models, when compared with control groups, G-CSF resulted in obvious thickening of the endometrium along with increased expression levels of cytokeratin and vimentin.The intrauterine perfusion of G-CSF, as a chemical and mechanical stimulus, is considered to induce the secretion of endogenous cytokines and activate endocrineparacrine pathways, thus leading to successful embryo implantation and pregnancy [15].At the genetic level, G-CSF can downregulate the expression of hsa_circ_0001550, the downstream target genes of hsa_circ_0001550 participate in regulating endometrial receptivity and embryo implantation, so G-CSF may affect embryo implantation by adjusting hsa_circ_0001550-miRNAmRNA interaction network [28].Furthermore, by promoting embryonic adhesion, cell migration, tissue remodeling, and the expression of angiogenesis-related genes, G-CSF may contribute to embryo implantation [29].G-CSF has been shown to be involved in embryonic growth and development [30] and play a role in the maintenance of pregnancy [31].
Several clinical studies have investigated the use of G-CSF to treat women with a thin endometrium with G-CSF.In 2011, Gleicher et al. [30] were the first to report four infertile women with a thin endometrium who successfully became pregnant by ART following the intrauterine perfusion of G-CSF.Subsequently, the same research group recruited another cohort of patients with a persistent thin endometrium who had failed conventional treatment and underwent intrauterine G-CSF perfusion.These researchers found that these patients achieved a low but reasonable clinical pregnancy rate following the intrauterine infusion of G-CSF [32].Since then, several studies have investigated the use of G-CSF to treat women with a thin endometrium.However, the results of these studies were inconsistent.For example, Mishra et al. [33] reported that G-CSF led to a slight increase in endometrial thickness in women with a persistent thin endometrium; however, G-CSF did not improve pregnancy rate in such patients.In another study, Swati et al. [34] found that G-CSF improved endometrial thickness only in about a third of women with a persistent thin endometrium.Kim et al. [35] reported that G-CSF improved the thickness of a thin endometrium without causing intrauterine adhesions and increased the chances of conception and embryo implantation.Due to such inconsistency, it was evident that we needed to perform a meta-analysis of the published studies to determine whether G-CSF can affect endometrial thickness and pregnancy.
In total, we investigated 8 studies (consisting of 6 RCTs and 2 non-RCTs including 673 patients) in this metaanalysis.We discovered that intrauterine infusion or the subcutaneous injection of G-CSF significantly increased embryo implantation and clinical pregnancy rates when compared with controls.The effect of G-CSF on clinical pregnancy rate and embryo implantation rate were consistent before and after subgroup analysis, irrespective of whether we were considering RCTs or not.The use of G-CSF did not have a significant effect on endometrial thickness.Three studies were used to analyze endometrial thickness, including two RCTs and one non-RCT.Based on the results of sub-group analysis for two RCT studies, we found that G-CSF did increase endometrial thickness.As we know women with thin endometrium may have a higher risk of preterm birth, low birth weight, and miscarriage, embryo transplantation in patients with thin endometrium should be cautious, delaying transplantation to the cycle of thickening before transplantation can be suggested [36].Therefore, G-CSF may improve the pregnancy outcome by increasing the thickness of endometrium.However, only two RCTs were involved in this analysis; consequently, there is a significant need for more RCTs to fully validate our findings.A systematic review and meta-analysis of the use of G-CSF in ART was previously published by Kamath et al. [37] in 2017; however, this study only included 1 non-RCTs and 3 RCTs with low quality evidence.In the present metaanalysis, we included 8 studies, including 6 RCTs and 2 non-RCTs; there has been an increase in relevant research over recent years and the level of evidence for the studies included in our study was better than the previous study [37].These previous authors reported significantly a higher clinical pregnancy rate after the use of G-CSF when compared with controls [RR = 2.43, 95% CI (1.09, 5.40), I 2 = 0%] but did not identify a statistically significant increase in endometrial thickness when G-CSF was used to treat women with a thin endometrium [MD = 0.47, 95% CI (1.36, 2.31), I 2 = 82%] [37].These conclusions are similar with our present findings and another meta-analysis conducted by Zhao et al. [38].However, the study of Zhao et al. [38] analyzed collated data from studies with prospective and retrospective designs; this may have affected analytical efficiency.
There are several limitations to our current analysis that need to be considered.First, the sample size of our analysis was small, as only five RCTs and two non-RCTs were included; this may have led to bias.Despite the limited number of studies, when stratified by RCTs vs. non-RCTs, subgroup analyses related to embryo implantation rate and clinical pregnancy rate showed consistent results with those of the overall pooled analyses.Inconsistent results were detected with regards to the influence of G-CSF on endometrium thickness; therefore, further studies are now needed to clarify the true benefits of G-CSF treatment.Second, the dosage and methods of G-CSF administration, as well as the underlying causes of thin endometrium, tended to vary across the included studies.Since the available studies did not individualize treatment strategies according to different underlying causes of thin endometrium, it was impossible to generate a definitive conclusion with this dataset.Therefore, well-designed randomized controlled studies with a large sample size and high quality are still needed to validate our findings.Current studies are still in the exploratory stage and there is no uniform standard for dosage, the times of administration, route of administration and timing of administration.

Conclusions
The results of our meta-analysis indicate that G-CSF may increase embryo implantation rate and clinical pregnancy rate of patients with thin endometrium during ART, G-CSF may have the potential to increase the endometrial thickness, although further studies are needed to fully validate the impact of G-CSF on endometrial thickness.Nevertheless, our findings should be considered cautiously conclusion owing to the limited number and quality of the studies involved in our meta-analysis, as well as differences in the etiology of thin endometrium, the frequency and route of G-CSF administration.Large-scale, high-quality, multicenter RCTs are now needed to fully elucidate the benefits of G-CSF on thin endometrium.

Fig. 2 .
Fig. 2. Assessment of risk of bias for the included RCTs.

Fig. 5 .
Fig. 5. Comparison of clinical pregnancy rate between G-CSF group and control group.

Fig. 6 .Fig. 7 .
Fig. 6.Comparison of clinical pregnancy rate between G-CSF group and control group in RCTs.

Fig. 8 .
Fig. 8.Comparison of endometrial thickness between G-CSF group and control group in RCTs.

Table footnotes
: The items are scored 0 (not reported), 1 (reported but inadequate) or 2 (reported and adequate).