Pegfilgrastim on febrile neutropenia in pediatric and adolescent cancer patients: a systematic review and meta-analysis

ABSTRACT Objectives There is no meta-analysis about the effects of pegfilgrastim on the occurrence of febrile neutropenia (FN) in pediatric/adolescent cancer patients. The study explored the efficacy of prophylactic pegfilgrastim in preventing FN in children/adolescents with cancer. Methods PubMed, Embase, and the Cochrane Library were searched for studies published before April 7, 2020. The primary outcome was the rate of FN. Effect size (ES) and odds ratio (OR) with 95% confidence intervals (CIs) were used to evaluate the outcome. The ES represented the rate of FN, and the STATA ‘metaprop’ command was used to synthesize the rate. Results Eight studies were included, comprising 167 patients and 550 courses of treatment. There was no difference between pegfilgrastim and filgrastim for the rate of FN in children receiving chemotherapy (OR = 0.68, 95% CI: 0.20–2.23, P = 0.520). In patients receiving pegfilgrastim, the rate of FN was 25.6% (95% CI: 14.9%−36.3%), the rate of grade 4 FN was 38.3% (95% CI: 19.2%−59.5%), the rate of severe neutropenia (SN) was 40.5% (95% CI: 35.1%−46.1%), and the rate of treatment delays due to FN was 4.8% (95% CI: 0.8%−11.3%). Discussion The number of studies that could be included was small; therefore, a specific type of cancer or a specific treatment could be studied. Heterogeneity was high. Conclusion There was no difference between pegfilgrastim and filgrastim for the rate of FN. The use of pegfilgrastim was still associated with rates of FN, grade 4 FN, severe neutropenia, and treatment delays due to FN in pediatric cancer patients.

The current guidelines recommend granulocyte colony-stimulating factors (G-CSF) for primary prophylaxis of chemotherapy-induced FN [3,6].G-CSF can also be used in the management of invasive fungal infections [7,8].Most available G-CSF treatments are shortacting, but new long-acting agents (e.g.pegfilgrastim and lipegfilgrastim) are being made available [9,10].Of note, short-acting G-CSF can exert their effect within 24-48 h, while long-acting G-CSF can take up to 1-2 weeks [9,10].Prophylaxis with G-CSF has been shown to reduce the incidence of FN and improve cancer treatment outcomes [11,12].Unfortunately, the use of G-CSF in clinical practice is suboptimal [13][14][15][16].Additional studies indicating the usefulness of G-CSF might improve this outcome.
Pegfilgrastim had a favourable efficacy and safety profile in clinical trials [17][18][19][20][21]. Meta-analyses showed that G-CSF in adult patients receiving chemotherapy is effective and safe, including pegfilgrastim and biosimilars [22], that it reduced the occurrence of FN compared with filgrastim [23], and that long-acting G-CSF had better outcomes compared with short-acting ones [24].Using filgrastim or pegfilgrastim vs. placebo improved the overall survival (OS) of patients receiving chemotherapy [25].
Nevertheless, there is only one guideline for pediatric and adolescent cancer patients regarding the management of FN [4], and there is no meta-analysis about the effects of pegfilgrastim on the occurrence of FN in pediatric and adolescent cancer patients.In addition, there is insufficient evidence to suggest that short-or longacting G-CSF can increase the survival of pediatric patients in FN.Moreover, G-CSF might have an economical cost and some potential risks, such as the risk of inducing acute myeloid leukemia or myelodysplastic syndromes [26].Therefore, this meta-analysis aimed to explore the efficacy of prophylactic pegfilgrastim in preventing FN in children/adolescents with cancer compared with G-CSF or no treatment/placebo.

Literature review
This meta-analysis was performed and reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) reporting guidelines [27].Since no original clinical raw data was collected or used, ethical approval was not requested for this meta-analysis.
PubMed, Embase, and the Cochrane Library were searched for studies published before April 7, 2020, using the MeSH terms 'child', 'pediatrics', 'adolescent', and 'pegfilgrastim', combined with relevant key words.
The exclusion criteria were: 1) conference abstract, case report, meta or review, animal study, and protocol; 2) language not English; 3) full text could not be obtained or no data available.

The primary outcome and any secondary outcomes
The primary outcome was the rate of FN.The secondary outcomes include rates of FN for pegfilgrastim vs. filgrastim, rate of grade 4 FN, rate of SN, and treatment delays.

Data extraction and quality assessment
Potentially relevant publications were screened and evaluated by two reviewers (Weiling Zhang and Yi Zhang) double-blindly, with a third reviewer (Huimin Hu) being requested to solve any disagreement.A structured data collection form was developed.Two researchers (Xia Zhu and Yizhuo Wang) independently extracted the data, including authors, year of publication, country, study design, sample size, age, percentage of males, episodes or courses, disease, dosage, etc.
The randomized controlled trials (RCTs) were evaluated using the Cochrane risk bias tool [28,29].The observational studies were evaluated using the Newcastle-Ottawa scale (NOS) [30].The case series studies were evaluated according to the National Institute for Health and Clinical Excellence (NICE) method [31].The non-randomized trials were evaluated according to the revised and validated version of the methodological index for non-randomized studies (MINORS) [32].Finally, the crossover studies were evaluated according to Hong Ding's quality assessment standard [33].

Statistical analysis
All analyses were performed using STATA MP 14.0 (Sta-taCorp, College Station, Texas, USA).Effect size (ES) and odds ratio (OR), and their 95% confidence intervals (CI) were used for analysis.The STATA 'metaprop' command was used to pool the rates.Statistical heterogeneity among the studies was calculated using Cochran's Q-test and the I2 index.An I 2 > 50% and Q-test P < 0.10 indicated high heterogeneity, and the random-effects model was used when high heterogeneity was present among studies; otherwise, the fixed-effects model was applied.P-values <0.05 were considered statistically different.Sensitivity analysis was performed to check the stability of the occurrence of FN for pegfilgrastim vs. filgrastim.Potential publication bias, Egger's test, and Begg's test were not performed due to the small number of studies included in each analysis [28].

Study selection
Figure 1 presents the study selection process.A total of 369 records were identified, and 303 were left after removing the duplicates.Then, 226 were excluded after screening, and 77 full-text papers were assessed for eligibility.Among them, 69 were excluded because of study aim/design (n = 25), population (n = 33), exposures (n = 6), data previously analyzed (n = 1), and non-English text (n = 4).
Finally, eight studies were included.There were four prospective studies [34][35][36][37] and four retrospective studies (Supplementary Table S1) [38][39][40][41].There were 167 patients and 550 episodes/courses.Four studies had no control group, and four studies used filgrastim as a control.Supplementary Table S2 presents the quality evaluation of the included study.

Sensitivity analysis
When considering the occurrence of FN for pegfilgrastim vs. filgrastim, no single study influenced the results (Supplementary Figure S1).

Discussion
The results suggest that compared with filgrastim, pegfilgrastim had a similar rate of FN.When using pegfilgrastim, the rate of FN was 25.6%, the rate of grade 4 FN was 38.3%, the rate of SN was 40.5%, and the rate of treatment delays due to FN was 4.8%.
In adults, studies showed that long-acting G-CSF achieves better outcomes than short-acting G-CSF [17][18][19][20][21], and meta-analyses in adults also indicated the superiority of long-acting G-CSF [22][23][24][25].Nevertheless, only one guideline [4] and no meta-analysis is currently available specifically for the management of FN in pediatric patients, and the guidelines highlight the  need for additional evidence [4].The results in children and adolescents in the present meta-analysis are supported by the results in adults [22][23][24][25].Nevertheless, analyses in adults showed that pegfilgrastim prevented treatment delays and dose reductions [22][23][24][25] since the neutrophil values have a higher likelihood of being within the adequate ranges, but this could not be observed in the present meta-analysis because of the available data.The FN rate with pegfilgrastim could be compared with filgrastim, and there was no statistically significant difference.In addition, the rates of FN, grade 4 FN, SN, and treatment delays were high (25.6%,38.3%, 40.5%, and 4.8%, respectively), and it is surprising since chemotherapyinduced SN and FN are considered the first reason for dose-limiting toxicity in pediatric cancer [34], but it has been reported that only small numbers of cycles were started later in pediatric patients treated with G-CSF, as the incidence of grade 4 SN and FN which may result in intolerance to the treatment was relatively low, reportedly about 6% [34,38].A previous meta-analysis showed that filgrastim reduced the rate of FN by 20% and shortened hospitalization time, but with no impact on infection-related mortality [42].The lack of difference in the present study between pegfilgrastim and filgrastim could be because G-CSF stimulates the hematopoietic cells to produce more neutrophils.The hematopoietic system in children is more active than the adult counterpart, and the composition of hematopoietic stem cells and their location in the body are also different [43,44].Therefore, future study designs should consider the stratification of children and adults to explore the effects of G-CSF.Furthermore, different chemotherapy drugs have different myelosuppression potentials [45,46], but they could not be taken into account in the present meta-analysis because of the small number of included studies.
The conclusions of this meta-analysis must be considered within its limitations.The number of studies that could be included was small, leading to a small number of patients and courses of treatment.Because of this, no selection was based on the type of cancer or chemotherapy, introducing bias due to the different aggressiveness of the diseases and treatments.If a control group was included, the control group was filgrastim or a placebo.The studies were conducted in various countries with different guidelines for using G-CSF.These reasons led to high heterogeneity in all analyses.Furthermore, the safety issues could not be analyzed because of the too important differences in safety outcomes among the included studies.In addition, future studies should examine the effect of pegfilgrastim biosimilars in pediatric patients with cancer.
In conclusion, the results suggest no difference between pegfilgrastim and filgrastim for the rate of FN.The rates of FN, grade 4 FN, SN, and treatment delays were 25.6%, 38.3%, 40.5%, and 4.8%, respectively, in pediatric patients receiving chemotherapy.However, high-level evidence is needed to define the role of pegfilgrastim in pediatric/adolescent cancer patients.

Figure 5 .
Figure 5. Forest plot of pooled severe neutropenia rate.