Continuous infusion of granulocyte colony-stimulating factor is associated with an advantage in neutrophil recovery in pediatric oncologic disorders

Abstract

Introduction

The use of intensified chemotherapy has significantly improved overall survival (OS) and relapse-free survival (RFS) in cancer patients. However, it also increases the risk of chemotherapy-induced neutropenia (CIN). Granulocyte colony-stimulating factor (G-CSF) is commonly used to hasten neutrophil recovery, but limited research has focused on the superiority of different routes of G-CSF administration.

Materials and methods

This was a randomized, prospective, single-institution study. Twenty-eight participants were enrolled and randomly assigned to experimental and control arms in a 1:1 ratio. At the end of the study, 20 cases were eligible for statistical analysis. The primary endpoint was the duration from the onset of neutropenia to steady neutrophil recovery. Secondary endpoints included the safety profile, 7-day emergency return rate, and the occurrence of infectious or febrile events.

Results

The mean duration from CIN to neutrophil recovery was 6.2 days in the intravenous drip (IVD) group and 7.6 days in the intravenous injection (IVI) group (P = 0.0068). The mean difference between the two groups was 1.4 days. A significantly lower incidence of febrile neutropenia (FN) was observed in the IVD group compared to the IVI group (15% vs. 50%, P = 0.0407). Adverse effects (AEs) were comparable between the two groups.

Conclusions

The IVD route of G-CSF administration leads to faster neutrophil recovery and reduces the risk of FN. All evaluable AEs are comparable between the two groups. These findings suggest that the IVD route may be a preferable option for optimizing supportive care in cancer patients with CIN.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s00280-026-04966-x.

Introduction

Intensified chemotherapy has improved relapse-free survival (RFS) and overall survival (OS) probabilities, albeit with the risk of treatment-related complications, including organ toxicity and mortality [1–5]. Advances in supportive care have reduced the incidence of treatment-related complications. Intensified chemotherapeutic agents result in significant myelosuppression, increasing the risk of infectious events such as sepsis or viral reactivation [6–10]. The occurrence of post-chemotherapy infectious events is associated with a poor survival prognosis [9, 11–13]. Clinical physicians vigorously strive to overcome treatment-related complications.

Among the various cell line declines secondary to chemotherapy-related marrow suppression, neutropenia remains the most challenging to manage. Anemia and thrombocytopenia can be managed short-term with blood product transfusions, lowering the risk of life-threatening events from anemia or thrombocytopenia. However, we rarely transfuse white blood cells to support neutropenia unless severe sepsis occurs. Instead, we turn to the use of granulocyte colony-stimulating factor (G-CSF) to accelerate the recovery phase of neutropenia.

G-CSF is a growth factor that stimulates progenitor cells in the marrow to hasten neutrophil recovery. The U.S. Food and Drug Administration (FDA) has approved subcutaneous injection, intravenous injection (IVI), and intravenous drip infusion (IVD) as routes of administration [14–19]. To prevent excessive skin puncture wounds, subcutaneous application is often avoided in pediatric hematology and oncology populations to reduce the risk of local infection. However, pharmacologically speaking, slower absorption routes, such as the subcutaneous one, are proven to be more effective in neutrophil recovery [15, 20–22]. This prospective study aims to determine if IVD G-CSF is superior to IVI in terms of neutrophil recovery after chemotherapy.

Materials & methods

Study objectives

We hypothesized that, in terms of G-CSF administration, the route of IVD would lead to faster neutrophil recovery compared to IVI.

Study design

We conducted this randomized, prospective, single-institution study to investigate our objectives. The study was approved by a national clinical trial (identifier: NCT06145321) and by the Chang Gung Memorial Hospital Institutional Review Board (IRB identifier: 202301323A3). The CONSORT checklist was provided in the supplementary file (S1). Patients with an absolute neutrophil count (ANC) lower than 500 cells/µL after chemotherapy were randomly assigned in a 1:1 ratio to either the experimental or control arms. The experimental arm received IVD G-CSF for 4 h at a dose of 5 mcg/kg, while the control arm received IVI G-CSF at the same dose. The G-CSF preparations used in the enrolled cohort were filgrastim and lenograstim. The IVD infusion solution was prepared using a weight-based, fixed dose of G-CSF diluted with 0.9% sodium chloride or 5% dextrose in water to a total volume of 20 mL. In our institutional practice, 0.9% sodium chloride was used as the diluent for lenograstim, whereas 5% dextrose in water was used for filgrastim. The infusion rate was set as 5 mL/hr to complete the administration over 4 h. Separately, IVI G-CSF bolus injection was administered over less than one minute. Three days after treatment initiation, complete blood counts, including differential counts, were monitored daily to determine the timing of steady neutrophil recovery. The experimental arm would switch to the control arm (and vice versa) to receive the corresponding G-CSF administration according to the study design if post-chemotherapy neutropenia event appears in the following hospitalization. This prospective clinical trial was closed in November 2024.

Eligibility criteria

Inclusion criteria

Pediatric patients aged between 0 and 18 years with a diagnosis of hematologic or oncologic malignancies were included in this clinical trial. Enrolled patients with an ANC lower than 500 cells/µL after chemotherapy were randomized using sealed envelopes in accordance with the study design.

Exclusion criteria

Patients were excluded if they had a diagnosis of myelodysplastic syndrome, aplastic anemia, or leukemia; had received G-CSF treatment within 7 days before enrollment; or were concurrently receiving cytokine therapy or thrombopoietin receptor agonist therapy.

Primary endpoint

The primary endpoint was the duration from the onset of chemotherapy-induced neutropenia (CIN) to steady neutrophil recovery, defined as an ANC higher than 1500 cells/µL consistently, a widely accepted clinical threshold indicating the resolution of neutropenia. This level signifies that the patient’s immune function has sufficiently recovered to safely defend against opportunistic pathogens.

Secondary endpoint

Secondary endpoints included the 7-day emergency return events, G-CSF-related adverse effects (AEs), occurrences of febrile neutropenia (FN), and sepsis. The AEs were evaluated using the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0. These were documented and analyzed.

Sample size calculation

The primary outcome is the number of days to neutrophil recovery, measured as continuous variable. Based on clinical relevance, an effect size of 0.8 was considered meaningful. Assuming a standard deviation of 1.0, with a power of 80% and a two-sided alpha of 0.05, the required sample size was calculated to be 25 participants per group. To account for a 10% dropout rate, the initial target enrollment was adjusted to 28 participants per group.

Statistical analysis

Mean values with ranges were used to describe quantitative variables while the frequency and percentage were used for categorical variables. Continuous variables, including the duration of neutrophil recovery, hospitalization duration, and blood cell counts, were analyzed using paired t-tests. Chi-square tests were performed to determine the statistical difference among categorical variables. All statistical tests were conducted using GraphPad Prism version 9.0 software. A p-value of less than 0.05 was considered statistically significant.

Results

Case enrollment and randomization

This trial was registered with the National Clinical Trial Registry under the identifier NCT06145321 on 24/11/2023. The first case was enrolled on 26/11/2023, and the last enrollment occurred on 07/10/2024. The trial concluded as scheduled on 30/11/2024. Twenty-eight pediatric patients met the inclusion criteria. Formal informed consent was obtained from the legal guardians of all children before study initiation. The enrolled population was categorized into an experimental arm (IVD group) and a control arm (IVI group) using a 1:1 randomization process. The treatment flowchart was shown in Fig. 1. Among the enrolled patients, 2 in the experimental arm dropped out early due to personal reasons. Nine patients (64%) in the experimental arm switched to the control arm to receive the assigned G-CSF therapy; 11 patients (79%) in the control arm switched to the experimental arm as per the study design, resulting in a total of 20 patients available for statistical analysis. Six patients across the IVD and IVI groups did not develop subsequent neutropenia meeting the protocol criteria, and therefore did not qualify for crossover phase. While their baseline demographics are provided, they were omitted from the final comparative analysis between the IVI and IVD arms.

Fig. 1.

Flowchart of case identification and enrollment

Characteristics of the study population

The baseline characteristics of the eligible population are summarized in Table 1. The study included 17 boys and 11 girls, with a mean age of 6.2 years (ranges, 1.5–18.0 years). Cancer subtypes included rhabdomyosarcoma (6 patients), neuroblastoma (7 patients), brain tumor (7 patients), hepatoblastoma (2 patients), Ewing sarcoma (2 patients), osteosarcoma (2 patients), retinoblastoma (1 patient), and extracranial germ cell tumor (1 patient). Nearly half of the patients (43%) had refractory or relapsed disease at enrollment. In the enrolled cohort, the majority of patients received filgrastim as the designated G-CSF treatment. Only one initially enrolled patient received lenograstim, and this patient did not complete the crossover treatment and was therefore not included in the final comparative analysis. A comparison of complete blood counts at study enrollment between the experimental and control groups is listed in Table 2. The complete blood count parameters were comparable across subgroups, with no statistically significant differences identified between the two study arms at enrollment.

Table 1.

Baseline characteristics of the enrolled study cohort

	Enrolled cases (n = 28)
Age (years)	6.2 (1.5–18.0)
Gender
Male	17 (61%)
Female	11 (39%)
Diagnosis
CNS tumor	7 (25%)
Neuroblastoma	7 (25%)
Rhabdomyosarcoma	6 (21%)
Hepatoblastoma	2 (7%)
Retinoblastoma	1 (4%)
Extracranial germ cell tumor	1 (4%)
Osteosarcoma	2 (7%)
Ewing sarcoma	2 (7%)
Disease status
R/R	12 (43%)
Non R/R	16 (57%)
G-CSF preparation
Filgrastim	27 (96%)
Lenograstim	1 (4%)

CNS, central nervous system; G-CSF, granulocyte colony-stimulating factor; R/R, refractory or relapsed

Table 2.

Comparisons between the IVD and IVI groups at study initiation

	IVD group (n = 20)	IVI group (n = 20)	P value
Age (years)	5.7 ± 3.2	5.6 ± 3.2	0.7204
Timing of initiation (post-chemotherapy day)	6.0 ± 1.5	5.8 ± 1.6	0.4283
WBC (/µL)	875 ± 492	950 ± 343	0.5274
ANC (/µL)	413 ± 221	481 ± 252	0.4342
Hb (g/dL)	8.7 ± 1.1	8.8 ± 1.4	0.7827
PLT (/µL)	139,750 ± 93,339	162,400 ± 108,840	0.2655

Data were presented as mean ± standard deviation unless otherwise specified

ANC, absolute neutrophil count; Hb, hemoglobin; IVD, intravenous drip; IVI, intravenous injection; PLT, platelet; WBC, while blood cell

Primary outcome

Efficacy and treatment responses between the IVD and IVI groups are presented in Table 3. The mean duration from CIN to neutrophil recovery was 6.2 days in the IVD group and 7.6 days in the IVI group (P = 0.0068), with a mean difference of 1.4 days between the two groups. The administration of IVD G-CSF results in neutrophil recovery that is 1.4 days faster compared to the IVI group. This statistically significant difference indicated that IVD administration was superior to IVI in promoting neutrophil recovery following CIN. Complete blood counts at the time of G-CSF cessation were recorded according to the study protocol. The IVD group exhibited a higher white blood cell count (4895 vs. 3665, P = 0.1155) and ANC (3253 vs. 2394, P = 0.1240), although statistical significance was not reached.

Table 3.

Comparisons between the IVD and IVI groups following the administration of G-CSF

	IVD group (n = 20)	IVI group (n = 20)	P value
Complete blood counts
WBC (/µL)	4895 ± 3612	3665 ± 1235	0.1155
ANC (/µL)	3253 ± 2212	2394 ± 871	0.1240
Hb (g/dL)	8.6 ± 0.9	9.0 ± 1.3	0.3084
PLT (/µL)	88,750 ± 85,917	98,350 ± 118,155	0.6988
AEs
Any grade	55%	65%	0.7475
Grade 1–2	55%	65%	0.7475
Grade 3–4	0	0	NA
Event of FN	15%	50%	0.0407*
Microbial-proven sepsis	0	0	NA
7-day ER return rate	0	0	NA
Time to neutrophil recovery	6.2 ± 3.7	7.6 ± 3.8	0.0068*

Data were presented as mean ± standard deviation unless otherwise specified

AEs, adverse events; ANC, absolute neutrophil count; ER, emergency room; FN, febrile neutropenia; Hb, hemoglobin; IVD, intravenous drip; IVI, intravenous injection; NA, not available; PLT, platelet; WBC, while blood cell; * P value < 0.05

Secondary outcome

G-CSF-related AEs were monitored and documented following the CTCAE grading system. Common AEs, such as bone pain, flu-like symptoms, and allergic reactions, were comparable between the two study groups (Table 3; Fig. 2). No grade 3 or 4 adverse effects were observed (Table 4). A significantly higher incidence of FN was observed in the IVI group compared to the IVD group (50% vs. 15%, P = 0.0407). No patients returned to the emergency room within 7 days due to discomfort or illness, resulting in a 7-day return rate of 0% in both the experimental and control groups. Other complete blood count parameters, such as hemoglobin and platelet counts, were also comparable between the IVD and IVI groups. This safety profile supported the conclusion that IVD administration is both safe and effective in managing CIN.

Fig. 2.

Common adverse events following the administration of G-CSF in the IVD and IVI groups

Table 4.

Various AEs based on CTCAE in the IVD and IVI groups

	IVD group (n = 20)	IVI group (n = 20)	P value
Bone pain
Grade 1	5%	10%	> 0.9999
Grade 2	5%	10%	> 0.9999
Grade 3–4	0	0	NA
Flu-like symptoms
Grade 1	20%	25%	> 0.9999
Grade 2	10%	20%	0.6614
Grade 3–4	0	0	NA
New-onset N/V
Grade 1	10%	15%	> 0.9999
Grade 2	0	0	NA
Grade 3–4	0	0	NA
Allergic reaction
Grade 1	5%	5%	> 0.9999
Grade 2	0	0	NA
Grade 3–4	0	0	NA

AEs, adverse events; CTCAE, common terminology criteria for adverse events; IVD, intravenous drip; IVI, intravenous injection; NA, not available; N/V, nausea or vomiting

Discussion

Over the past decades, the survival probability of cancer patients has improved significantly due to scientific progress. The introduction of multi-agent and intensified chemotherapy has improved RFS and prolonged OS [2, 4, 8]. The development of targeted therapy and immunotherapy has also assisted conventional cancer therapy in eliminating cancer growth. However, while multidisciplinary cancer management brings about significant survival benefits, it concomitantly leads to increased risks of treatment-related toxicity, which may influence the quality of life during surveillance [23–28].

Among the various AEs secondary to intensified cancer treatments, neutropenia remains the most common and troublesome. It is evident that the risk of infection correlates positively with the severity of neutropenia. Studies have shown that serious infection is an independent prognostic index in cancer populations [6, 7, 9–11]. The shorter the duration of neutropenia after chemotherapy, the lower the incidence of infection-related comorbidities such as sepsis.

G-CSF is an FDA-approved pharmacological agent for the management of CIN [14–16]. It stimulates the bone marrow, particularly the myeloid series, to hasten neutrophil recovery. The use of G-CSF significantly lowers the incidence of neutropenia-related sepsis and thus improves OS [29–32]. Although the clinical benefit of G-CSF application is well established, the optimal administration route is still debated. Theoretically, the subcutaneous route leads to more steady drug release, which is believed to be associated with a better treatment response [21, 22]. However, due to the risk of skin infection from excessive puncture and the fear of subcutaneous injection in pediatric patients, pediatric hematologists and oncologists often opt for the IVI or IVD route for G-CSF administration. Unfortunately, few studies focus on the superiority of one route of G-CSF administration over another. In our prospective trial, we have demonstrated that the IVD route is superior to the IVI route in terms of neutrophil recovery. The mean duration from CIN to neutrophil recovery was 6.2 days in the IVD group and 7.6 days in the IVI group (P = 0.0068). However, this benefit does not translate into a direct impact on the 7-day emergency return rate and the incidence of sepsis after intensified chemotherapy.

Numerous studies have highlighted the relationship between prolonged neutropenia and an increased risk of febrile and/or infectious events [33–37]. However, in our study, we were unable to draw the same conclusion. We believe that the relatively low incidence of microbiologically proven infectious events was the major contributing factor. With the improved quality of supportive care in the modern era and the early introduction of empirical antimicrobial agents, the incidence of microbiologically proven sepsis has decreased dramatically compared to that of past decades [38–41]. In our cohort, none of the enrolled cases in either the IVD or IVI groups experienced microbiologically proven sepsis. The incidence of FN in the IVD and IVI groups was 15% and 50%, respectively. A significantly lower occurrence of FN was identified in the IVD group, suggesting a potential association between reduced FN risk and the use of IVD G-CSF.

The AEs of G-CSF are dose-dependent. Higher doses of G-CSF lead to an increased incidence of treatment-related complications, including bone pain, allergic reactions, and febrile illness [42–45]. These conditions are more evident when G-CSF is used to mobilize progenitor cells for harvest [46]. On the other hand, to prevent the development of serious infection from neutropenia, clinical physicians may titrate up the therapeutic dose of G-CSF if tolerated, hoping to decrease the duration of neutropenia. Such clinical application may lead to faster neutrophil recovery but inevitably results in increased events of G-CSF-mediated discomforts. In this prospective study, possible G-CSF-related AEs were documented, evaluated, and analyzed to generate the safety profile. Incidences of bone pain and flu-like symptoms were 10% and 30% in the IVD group and 20% and 45% in the IVI group, respectively. All evaluable AEs after the application of G-CSF were comparable between the experimental and control groups. The safety profile in our study indicates that the IVD route does not exhibit increased risks of treatment-related toxicity while maintaining its effectiveness on neutrophil recovery.

Despite the promising benefits in the IVD subgroup, there are still some limitations to our study. First, although it is a prospective and single-blinded clinical trial, the relatively small population size and single-institution setting may introduce bias into the conclusions. Second, as a real-world practice study, the clinical management of G-CSF infusion is believed to increase the workload for the nursing staff, making their tasks more time-consuming. Confounding factors such as variations in care quality and differences in nursing staff skills cannot be entirely eliminated. Third, a portion of the enrolled patients failed to complete the intended G-CSF management due to early dropout from the study. Most of these patients were young children within the IVD subgroup, indicating that the inconvenience of continuous infusion remains an obstacle in clinical practice, despite the established clinical benefit of neutrophil recovery. Regarding the two different kinds of G-CSF preparations involved in the study, this remained a substantial confounding factor for the results. However, the only patient who used lenograstim did not complete the crossover. Consequently, all patients included in the crossover comparison received filgrastim, minimizing potential heterogeneity related to different G-CSF preparations in the primary analysis. As for the issue of a potential sequence or period effect, we agree that this is an important consideration in a crossover study. The present study was primarily designed to compare neutrophil recovery between the two G-CSF administration methods and was not adequately powered to formally evaluate treatment-by-period or sequence effects. Therefore, we cannot exclude the possibility that treatment sequence or differences between chemotherapy cycles influenced the observed neutrophil recovery. Finally, the relatively small final sample size available for the primary comparative analysis is another concern. Although our prospective trial successfully met its initial target of 28 patients, 6 patients did not experience a recurrent neutropenic event and therefore did not proceed to the crossover phase. 2 patients dropped out early. Consequently, only 20 patients were included in the final analysis. This sample attrition may have limited the statistical power to detect subtle differences between the IVI and IVD G-CSF administration. Hence, our findings remain exploratory and hypothesis-generating. Future multi-center studies with larger cohorts are warranted to validate our findings.

Conclusions

In our study, we demonstrated that IVD G-CSF administration accelerates neutrophil recovery by 1.4 days compared to the IVI route. Additionally, the incidence of FN was significantly lower in the IVD group. The risk of AEs was balanced between continuous infusion and intravenous injection. These findings support the potential clinical advantage of the IVD route for optimizing supportive care in cancer patients with CIN.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

AEs

Adverse effects

ANC

Absolute neutrophil count

CIN

Chemotherapy-induced neutropenia

CTCAE

Common Terminology Criteria for Adverse Events

FDA

Food and Drug Administration

FN

Febrile neutropenia

G-CSF

Granulocyte colony-stimulating factor

IQR

Interquartile ranges

IVD

Intravenous drip infusion

IVI

Intravenous injection

OS

Overall survival

RFS

Relapse-free survival

Author contributions

Conceptualization Tsung-Yen Chang; methodology and formal analysis Tsung-Yen Chang; writing—original draft preparation Yi-Lun Wang; writing—review and editing Yi-Lun Wang, Shih-Hsiang Chen and Tang-Her Jaing; visualization Yi-Wen Hsiao; project administration Yu-Chuan Wen; All authors have read and agreed to the published version of the manuscript.

Funding

This study received no specific grant from any funding agency.

Data availability

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

Declarations

Conflict of interest

The authors declare no competing interests.

Ethics, Consent to Participate, and Consent to Publish

Ethical approval, as well as consent to participate and publish, were obtained through formal informed consent from the participants’ legal guardians. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Chang Gung Memorial Hospital (IRB identifier: 202301323A3).