Primary G-CSF prophylaxis in sacituzumab govitecan–treated mTNBC: real-world evidence from a multinational CEBCC-102 cohort

Abstract

Background

Sacituzumab govitecan (SG) is a key treatment for metastatic triple-negative breast cancer (mTNBC). Neutropenia is a common toxicity, and the optimal use of G-CSF (primary prophylaxis versus reactive administration) remains uncertain. We evaluated the association of primary G-CSF prophylaxis with clinical outcomes and treatment-related toxicities.

Methods

We conducted a multinational retrospective real-world cohort study of 303 patients with mTNBC treated with SG in the second or a later line. Efficacy endpoints included progression-free survival (PFS) and overall survival (OS), analyzed using the Kaplan–Meier method. Adverse events were assessed according to CTCAE criteria.

Results

Of 303 patients, 100 were included in the primary prophylaxis group and 203 in the no primary prophylaxis group. Median PFS was 4.2 months in the primary prophylaxis group versus 5.1 months in the no primary prophylaxis group (p = 0.66), with 6-month PFS rates of 38.5% vs. 42.1%, respectively. Median OS was 10.9 vs. 11.6 months (p = 0.95), with 12-month OS rates of 44.9% vs. 47.1%. The frequencies of the recorded individual AEs were similar between groups, apart from grade 3–4 neutropenia, which was less frequent in the primary prophylaxis group (33.0% vs. 50.7%, p = 0.005). Febrile neutropenia occurred in 4.0% vs. 4.4% of patients (p = 1.00). Other recorded AEs (anemia, diarrhea, nausea, vomiting, alopecia) were comparable. Dose reductions due to toxicity were similar (36.0% vs. 38.4%, p = 0.71). In the no primary prophylaxis group, 74.4% subsequently received secondary G-CSF.

Conclusions

In this retrospective real-world cohort, primary G-CSF prophylaxis was associated with a lower frequency of grade 3–4 neutropenia. No statistically significant differences in PFS, OS, or febrile neutropenia were observed between groups. These findings support a risk-adapted rather than universal approach to primary G-CSF prophylaxis.

Introduction

Sacituzumab govitecan (SG) has significantly improved outcomes in patients with metastatic triple-negative breast cancer (mTNBC). In the phase 3 ASCENT trial, SG demonstrated superior progression-free survival (PFS) and overall survival (OS) compared with standard chemotherapy. Median PFS was 5.6 months in the primary analysis [1] and 4.8 months in the final analysis [2], while median OS was 12.1 and 11.8 months, respectively. Although SG has mainly been used after progression on first-line systemic therapy for mTNBC, its role may expand following the PFS benefit reported with SG combined with pembrolizumab in the first-line setting [3].

Neutropenia is one of the most clinically relevant toxicities of SG. In ASCENT, neutropenia was the most common treatment-related adverse event, occurring in 63% of patients, and was also the predominant grade 3–4 toxicity, reported in 51% of patients [1]. Grade 3 and grade 4 febrile neutropenia (FN) occurred in 5% and 1% of patients, respectively, while growth-factor support was administered to 49% of patients receiving SG [1].

Recommendations regarding primary granulocyte colony–stimulating factor (G-CSF) prophylaxis during SG therapy have recently evolved. The National Comprehensive Cancer Network (NCCN) Hematopoietic Growth Factors Guidelines, Version 1.2026, classified SG among regimens associated with a high risk of FN, exceeding 20% [4]. In contrast, Version 3.2026 classifies SG as an intermediate-risk regimen, with an estimated FN risk of 10–20% [5]. The Summary of Product Characteristics (SmPC) for SG recommends considering primary G-CSF prophylaxis in patients at increased risk of FN [6]. Therefore, G-CSF use should be individualized according to clinical risk factors, while secondary prophylaxis remains appropriate when neutropenia causes treatment delays in patients not receiving G-CSF from the first cycle.

Differences between outcomes observed in randomized clinical trials and those seen in routine practice are commonly described as the efficacy–effectiveness gap [7]. This issue is particularly relevant for toxicity and supportive-care strategies, as patients treated in real-world settings may be older, more comorbid, and less selected than clinical trial populations [8].

Therefore, this study aimed to evaluate the real-world use of primary G-CSF prophylaxis and to assess its associations with efficacy and safety outcomes by comparing an upfront primary prophylaxis strategy with a predominantly reactive strategy in a large multinational cohort of patients with mTNBC treated with SG.

Methods

Data source and study design

This retrospective, multicenter cohort analysis was conducted across 18 oncology centers in the Czech Republic, Poland, and Slovakia. The study included consecutive patients with metastatic triple-negative breast cancer (mTNBC) who received sacituzumab govitecan (SG) in routine clinical practice in the second or a later line of treatment for advanced disease. The investigation was performed within the framework of the Central European Breast Cancer Collaboration (CEBCC)−102 initiative [9]. Patients included in the study initiated SG treatment between August 2021 and April 2025. The data cut-off for follow-up and analysis was July 1, 2025.

The present study is a secondary analysis of the previously reported 303-patient CEBCC-102 cohort [9], focused specifically on G-CSF prophylaxis. A related age-specific analysis of the CEBCC-102 cohort and an earlier Polish real-world SG cohort have been reported separately [10, 11].

Objectives and endpoints

The primary endpoint was progression-free survival (PFS). Secondary endpoints included overall survival (OS) and recorded safety outcomes, including any-grade neutropenia, grade 3–4 neutropenia, febrile neutropenia, other recorded adverse events (AEs), and SG dose reductions due to AEs. Associations between primary G-CSF prophylaxis and these outcomes were evaluated.

PFS was defined as the time from initiation of SG to documented disease progression or death due to any cause. OS was defined as the time from initiation of SG to death from any cause. Adverse events (AEs) were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI-CTCAE), version 5.0 [12]. Efficacy assessments were performed using Response Evaluation Criteria in Solid Tumors (RECIST) v1.1 [13].

Safety analysis

Treatment-related adverse events were identified retrospectively from electronic medical records over the entire SG treatment period and graded according to CTCAE version 5.0. When an adverse event occurred more than once, the highest recorded grade per patient was used. Each patient was counted once for each adverse event; patients without documentation of a given event were classified as not having experienced it.

Primary and secondary G-CSF prophylaxis definitions

Administration of G-CSF from the first SG cycle, before neutropenia onset, was defined as primary prophylaxis, whereas initiation following a neutropenic episode was defined as secondary prophylaxis.

Statistical analysis

Descriptive statistics were used to summarize the data. Continuous variables were reported as mean (standard deviation) or median [interquartile range (IQR)]. Categorical variables were summarized as number and proportion of the total study population and by subgroups where appropriate. Differences between groups were evaluated using the Fisher exact test for categorical variables and the Wilcoxon rank-sum test for continuous variables. Time-to-event endpoints (PFS and OS) were estimated using the Kaplan–Meier method, with differences between groups compared using the log-rank test. Patients with insufficient data to calculate a given time-to-event endpoint were excluded only from the analysis of that endpoint. All statistical tests were two-sided, and p < 0.05 was considered statistically significant. Analyses were performed using Stata software (version 18; StataCorp, College Station, TX, USA).

Ethics considerations

The study was approved by the ethics committees of the Maria Sklodowska-Curie National Research Institute of Oncology in Krakow (approval no. 2/2023; 18 April 2023) and Warsaw (approval no. 21/2024; 22 February 2024), with additional approvals from the institutional review boards of the Masaryk Memorial Cancer Institute in Brno (approval no. 1737/2025; 10 June 2025) and the Ethics Committee of the General University Hospital in Prague (approval no. 110825 S-IV; 21 August 2025). The study was conducted in accordance with the Declaration of Helsinki and applicable national regulations. Due to the retrospective design and use of de-identified data, written informed consent was waived at most participating institutions in line with local ethics policies.

Results

A total of 303 female patients were included in the study, with a mean age of 54.7 years (range, 26.6–86.1). One hundred patients (33.0%) received primary G-CSF prophylaxis (primary prophylaxis group), whereas 203 patients (67.0%) did not (no primary prophylaxis group). The recorded baseline characteristics were generally similar between groups, except for ECOG performance status. In the primary prophylaxis group, a higher proportion of patients had no signs of functional impairment (ECOG 0). The results are summarized in Table 1. The median follow-up was 8.9 months (range, 0.2–32.0 months).

Table 1.

Baseline characteristics of patients in the primary prophylaxis group and no primary prophylaxis group

Characteristics	Primary prophylaxis group (n = 100)	No primary prophylaxis group (n = 203)	p
Previous (neo)adjuvant systemic therapy	74 (74.0%)	163 (80.3%)	0.24
At least one recorded relevant chronic condition	43 (43.0%)	95 (46.8%)	0.54
ECOG 0	50 (50.0%)	75 (37.1%)	0.035
ECOG ≥ 1	50 (50.0%)	127 (62.9%)	0.035
Lung metastases	50 (50.0%)	111 (54.7%)	0.46
Liver metastases	34 (34.0%)	65 (32.0%)	0.80
CNS metastases	10 (10.0%)	19 (9.4%)	0.84
Bone metastases	37 (37.0%)	85 (41.9%)	0.46
One previous line of systemic therapy for metastatic disease	48 (48.0%)	111 (54.7%)	0.33
More than one previous line of systemic therapy for metastatic disease	52 (52.0%)	92 (45.3%)	0.33
Previous febrile neutropenia before SG initiation	4 (4.0%)	13 (6.4%)	0.60

Abbreviation: CNS, central nervous system; ECOG, Eastern Cooperative Oncology Group; SG, sacituzumab govitecan. ECOG performance status was unavailable for one patient in the no primary prophylaxis group. Percentages for ECOG categories were calculated among patients with available ECOG data

Treatment efficacy

The median PFS in the primary prophylaxis group was 4.2 months versus 5.1 months in the no primary prophylaxis group (p = 0.66). The 6-month PFS was 38.5% (95% CI, 28.5–48.5) in the primary prophylaxis group versus 42.1% (95% CI, 34.9–49.2) in the no primary prophylaxis group. The median OS in the primary prophylaxis group was 10.9 months versus 11.6 months in the no primary prophylaxis group (p = 0.95). The 12-month OS was 44.9% (95% CI, 33.9–55.3) in the primary prophylaxis group versus 47.1% (95% CI, 39.2–54.6) in the no primary prophylaxis group. Results are presented in Figs. 1 and 2, respectively. PFS was evaluable in 299 patients, including 100 patients in the G-CSF primary prophylaxis group and 199 in the no primary prophylaxis group. OS was evaluable in 301 patients, including 100 and 201 patients, respectively. Four patients were excluded from the PFS analysis and two from the OS analysis because of insufficient data to calculate the respective endpoints.

Fig. 1.

Kaplan–Meier curves for progression-free survival in the primary prophylaxis group (Primary G-CSF) and no primary prophylaxis group (No primary G-CSF). Most patients in the no primary prophylaxis group subsequently received G-CSF as secondary prophylaxis. Abbreviation: G-CSF, granulocyte colony-stimulating factor

Fig. 2.

Kaplan–Meier curves for overall survival in the primary prophylaxis group (Primary G-CSF) and no primary prophylaxis group (No primary G-CSF). Most patients in the no primary prophylaxis group subsequently received G-CSF as secondary prophylaxis. Abbreviation: G-CSF, granulocyte colony-stimulating factor

Treatment safety

The frequencies of the recorded individual AEs were similar between groups, apart from grade 3–4 neutropenia, which was more frequent in the no primary prophylaxis group. Febrile neutropenia occurred in 4.0% of patients in the primary prophylaxis group and 4.4% in the no primary prophylaxis group (p = 1.00). SG dose reductions due to AEs were also similar between groups. Results are presented in Table 2.

Table 2.

Recorded adverse events and SG dose reductions according to primary prophylaxis status

Characteristics	Primary prophylaxis group (n = 100)	No primary prophylaxis group (n = 203)	p
SG dose reduction due to AEs	36 (36.0%)	78 (38.4%)	0.71
Febrile neutropenia during SG	4 (4.0%)	9 (4.4%)	1.00
Any-grade neutropenia	62 (62.0%)	147 (72.4%)	0.09
Grade 3–4 neutropenia	33 (33.0%)	103 (50.7%)	0.005
Diarrhea	19 (19.0%)	40 (19.7%)	1.00
Thrombocytopenia	9 (9.0%)	21 (10.4%)	0.84
Anemia	35 (35.0%)	85 (41.9%)	0.26
Vomiting	11 (11.0%)	21 (10.4%)	0.85
Nausea	24 (24.0%)	48 (23.7%)	1.00
Fatigue	41 (41.0%)	75 (37.0%)	0.45
Alopecia	53 (53.0%)	91 (44.8%)	0.13

AE, adverse event; G-CSF, granulocyte colony-stimulating factor; SG, sacituzumab govitecan. Values are presented as n (%). Treatment-related adverse events were assessed over the entire SG treatment period using the highest recorded CTCAE grade per patient. Each patient was counted once for each adverse event. All patients were included in the denominators for all variables (n = 100 and n = 203 in the primary and no primary prophylaxis groups, respectively)

Secondary G-CSF prophylaxis

Among patients in the no primary prophylaxis group, 151 (74.4%) subsequently received G-CSF as secondary prophylaxis. Ultimately, only 52 patients (17.2%) did not receive G-CSF at any time.

Discussion

This multicenter real-world study provides clinically relevant evidence on G-CSF prophylaxis strategies in patients with mTNBC receiving SG. The key finding is that an upfront G-CSF strategy was associated with a lower rate of grade 3–4 neutropenia than a predominantly reactive strategy. No statistically significant differences in PFS, OS, or FN were observed. This observation supports a risk-adapted rather than universal approach to primary G-CSF prophylaxis. Importantly, the high rate of secondary G-CSF use in the no primary prophylaxis group indicates that the absence of a survival difference should not be interpreted as evidence against G-CSF support, but rather as a lack of demonstrable advantage of routine upfront prophylaxis over timely reactive use in this cohort.

According to the SmPC, SG may cause severe or life-threatening neutropenia, and fatal infections in the context of neutropenia have been reported, most often during the first two treatment cycles [6]. Therefore, primary G-CSF prophylaxis should be considered from the first cycle in patients at increased risk of FN, including older patients, particularly those aged ≥ 65 years, patients with a history of neutropenia, poor performance status, organ dysfunction, or multiple comorbidities [6]. This regulatory position is consistent with the recent evolution of the NCCN Hematopoietic Growth Factors Guidelines. SG was classified as a high-risk regimen for FN in Version 1.2026 but was reclassified as an intermediate-risk regimen in Version 3.2026 [4, 5]. For intermediate-risk regimens, G-CSF use should be guided by individual patient-related risk factors rather than administered uniformly to all patients. This is particularly relevant in later-line treatment, where prior chemotherapy exposure and accumulation of additional patient-specific risk factors may increase the individual FN risk beyond the regimen-related risk alone. Thus, treatment history should be considered when selecting patients for primary G-CSF prophylaxis. In our cohort, previous FN before SG initiation was uncommon and did not differ significantly between groups, suggesting that primary prophylaxis decisions may have been driven by broader clinical judgment and local practice patterns rather than by this single risk factor.

The higher proportion of patients with ECOG 0 in the primary prophylaxis group was counterintuitive, as poorer performance status is generally considered a risk factor supporting G-CSF prophylaxis. This pattern may reflect clinician- and center-specific supportive-care practices, including a preference to preserve planned treatment intensity in fitter patients. Importantly, because poorer performance status was more frequent in the no primary prophylaxis group, this imbalance represents a potential source of confounding and should be considered when interpreting the similar survival outcomes between groups.

The PRIMED trial provides the most direct prospective evidence for upfront supportive care during SG therapy. This open-label, single-arm phase II study evaluated primary prophylaxis with G-CSF and loperamide in patients with HER2-negative advanced breast cancer treated with SG [14]. During the first two treatment cycles, any-grade neutropenia and diarrhea occurred in 28% and 34% of patients, respectively, while grade 3–4 neutropenia was reported in 16%. Grade ≥ 2 diarrhea occurred in 16% of patients, including grade 3 diarrhea in 4%. AEs led to dose reductions in 18% and treatment interruptions in 44%, with no discontinuations due to toxicity [14]. Neutropenia rates were higher in our cohort than in PRIMED. This may reflect differences in patient selection and treatment setting, as our patients received SG in the second or a later line in routine clinical practice, with greater heterogeneity in prior treatment exposure and supportive-care approaches, whereas PRIMED prospectively applied a predefined prophylactic strategy in a selected trial population. In our cohort, despite the absence of routine primary loperamide prophylaxis, diarrhea rates were almost identical between patients who received primary G-CSF prophylaxis and those who did not. No significant differences were observed for the other recorded non-hematologic AEs, suggesting that the main measurable difference associated with primary G-CSF prophylaxis in this dataset was lower grade 3–4 neutropenia.

Real-world evidence provides important context for interpreting these findings. In a US real-world study including 230 patients, Kalinsky et al. reported that 58% of patients received G-CSF during SG treatment, although primary and secondary prophylaxis were not distinguished. Median OS was 10.0 months overall, with similar outcomes in patients who did and did not receive concomitant G-CSF [15]. In another US real-world analysis by Nanda et al., including 381 patients with mTNBC treated with SG in the second line or beyond, median OS was 11.3 months and median time to next treatment or death was 5.6 months. Any G-CSF use was recorded in 59% of patients, and neutropenia was less frequent among patients receiving any G-CSF prophylaxis than among those who did not [16]. An abstract-based retrospective TriNetX analysis also reported substantial real-world G-CSF use during SG treatment; however, its coded EHR design limits direct comparison of neutropenia rates with prospectively ascertained clinical trial data [17]. These observations are broadly concordant with our findings, in which primary G-CSF prophylaxis was associated with lower grade 3–4 neutropenia without observed survival differences.

Our previous multicenter analysis of premedication and G-CSF use in SG-treated patients with TNBC also showed substantial heterogeneity in supportive-care strategies across centers, reflecting evolving SmPC guidance, reimbursement conditions, and risk-adapted local practice [18]. The present study extends those observations by specifically separating primary from secondary G-CSF prophylaxis and by linking this distinction with efficacy and safety outcomes. This distinction is clinically important, as most available real-world studies report any G-CSF use without specifying whether it was administered from cycle 1 or introduced after neutropenia.

The efficacy outcomes observed in the present cohort are also consistent with the broader real-world SG literature. A recent single-arm meta-analysis comparing clinical trials and real-world studies of SG in mTNBC confirmed the expected efficacy–effectiveness gap, with numerically lower outcomes in real-world populations than in clinical trials [19]. Large contemporary real-world cohorts from Turkey, the UK, France, and Southern Spain have reported median PFS and OS values broadly within the range observed in our dataset [20–23]. These data support the external validity of SG effectiveness in routine practice, while also emphasizing the need for optimized toxicity management in less selected patient populations.

Few studies have specifically assessed primary versus secondary G-CSF prophylaxis. In our cohort, approximately one-third of patients received primary G-CSF prophylaxis, while 74.4% of patients initially managed without primary prophylaxis ultimately received secondary G-CSF. Overall, only 17.2% of patients did not receive G-CSF at any time during SG therapy. This proportion appears lower than in several other real-world reports, in which approximately 40–50% of patients did not receive G-CSF during SG treatment [15, 16, 24]. The single-center analysis by Tam et al. further highlights the clinical and economic burden of G-CSF use in SG-treated patients, reporting substantial G-CSF utilization and associated drug costs [24]. Although pharmacoeconomic outcomes were not evaluated in our study, these data suggest that rational selection of patients for primary prophylaxis may have practical relevance beyond toxicity prevention.

Two broad toxicity-management patterns can be identified in real-world SG practice: greater reliance on G-CSF support or greater reliance on SG dose reductions. In the study by Kalinsky et al., 58% of patients received G-CSF and 34% underwent SG dose reductions [15]. In the study by Nanda et al., the corresponding rates were 59% and 51% [16]. Alaklabi et al. reported G-CSF use in 47% and dose reductions in 51% of patients, while Reinisch et al. reported rates of 34% and 48%, respectively [25, 26]. In our cohort, more than four out of five patients received G-CSF at some point during therapy, whereas approximately one-third required SG dose reduction. This pattern suggests a supportive-care-oriented approach, with relatively liberal use of G-CSF and less reliance on dose reduction.

The relationship between dose modification and SG efficacy requires careful interpretation. In our study, dose reductions due to AEs did not differ significantly between the primary prophylaxis and no primary prophylaxis groups. Safety analyses from ASCENT support dose modifications as part of toxicity management during SG treatment [27]. However, exposure-response analyses indicate that higher SG exposure is associated with a greater probability of response and longer survival outcomes, supporting the approved 10 mg/kg dose as the standard starting dose [28]. In addition, a Polish real-world study showed that initial SG dose reductions of ≥ 20% were independently associated with shorter PFS and OS without reducing toxicity [29]. Therefore, while reactive dose modifications remain appropriate when clinically indicated, proactive dose reduction should not replace optimized supportive care in patients who can receive standard-dose SG.

The optimal timing and schedule of G-CSF administration with SG remain incompletely defined. In PRIMED, primary prophylaxis consisted of filgrastim on days 3, 4, 10, and 11 [14]. NCCN recommends administering filgrastim until post-nadir absolute neutrophil count recovery to near-normal levels and maintaining a minimum interval of 12 days between pegfilgrastim and the next chemotherapy cycle [5]. Practical recommendations for ADCs emphasize avoiding G-CSF administration within 24 h before or after cytotoxic treatment and tailoring short-acting or long-acting G-CSF to the day of neutropenia within the SG cycle [30]. However, NCCN also acknowledges that ADCs have a longer half-life than the unconjugated chemotherapy payload and that the optimal cycle day for myeloid growth factor administration has not been systematically established [5]. This uncertainty is particularly relevant for SG because of its day 1 and day 8 schedule.

Recent pharmacovigilance and meta-analytic data further support close early monitoring during SG treatment. Xu et al. identified a strong neutropenia reporting signal, with most FAERS reports occurring within the first 30 days, while their randomized-trial meta-analysis showed an increased risk of all-grade neutropenia with SG; however, FAERS disproportionality does not estimate incidence or establish causality [31].

An additional area for future refinement is pharmacogenetic risk stratification. The uridine diphosphate glucuronosyltransferase 1A1 (UGT1A1) pathway is involved in SN-38 metabolism, and some studies have associated the UGT1A1*28 polymorphism with increased SG-related toxicity. In a recent real-world study, patients homozygous for UGT1A1*28 had a significantly higher incidence of FN than patients with wild-type or heterozygous genotypes, although no clear effect on PFS or OS was observed [32]. By contrast, Xu et al. did not identify a statistically significant difference in neutropenia risk across the evaluated UGT1A1 genotypes [31]. UGT1A1 testing was not performed in our cohort and cannot be used to explain the observed toxicity patterns. Overall, evidence regarding UGT1A1-based prediction remains inconsistent, and its clinical utility for selecting patients for primary G-CSF prophylaxis has not been established.

The clinical relevance of optimized G-CSF use is likely to increase as SG moves into earlier lines of mTNBC treatment. ASCENT-03 demonstrated improved PFS with SG compared with chemotherapy in previously untreated advanced TNBC patients not eligible for PD-1 or PD-L1 inhibitors, with high-grade AEs remaining common [33]. In addition, ASCENT-04/KEYNOTE-D19 supports the role of SG combined with pembrolizumab in the first-line treatment of PD-L1-positive advanced TNBC [3]. Patients receiving SG earlier in the disease course may have less cumulative bone marrow toxicity and fewer treatment-related risk factors for FN than the later-line population evaluated in our study. Therefore, the present findings should be extrapolated cautiously to first-line SG use. At the same time, as SG exposure begins earlier and may continue for longer treatment durations, prevention and management of neutropenia will become increasingly important for preserving dose intensity, avoiding treatment delays, and maintaining patient quality of care.

This study has several limitations. Its retrospective, multicenter real-world design precludes causal inference and introduces potential selection bias and confounding by indication. The decision to initiate primary G-CSF prophylaxis was clinician driven and likely reflected individual risk assessment and local practice patterns. The imbalance in ECOG performance status between groups, with poorer performance status more frequent in the no primary prophylaxis group, may have influenced both prophylaxis use and clinical outcomes. We did not capture granular details of G-CSF administration, including exact timing within cycle 1, product type, dose, duration, adherence, or reasons for escalation to secondary prophylaxis. The high rate of secondary G-CSF use in the no primary prophylaxis group may have diluted differences attributable specifically to upfront prophylaxis. PFS was assessed in routine practice without a protocol-mandated imaging schedule or central radiology review. Safety data relied on medical record documentation across centers, which may have resulted in underreporting or heterogeneous CTCAE grading. The low number of FN events limited statistical power for this endpoint and precluded robust subgroup analyses. Finally, subsequent lines of therapy and other supportive-care measures were not standardized and could have influenced OS. These limitations support interpreting the findings as associative and hypothesis-generating and underscore the need for prospective studies with predefined supportive-care algorithms and standardized G-CSF schedules.

Conclusions

Our study represents one of the largest real-world evidence analyses to date on G-CSF use with SG. Primary G-CSF prophylaxis was associated with a lower rate of grade 3–4 neutropenia, while no statistically significant differences in PFS, OS, or FN were observed. Given that most patients in the no primary prophylaxis group subsequently received secondary G-CSF, these findings support a risk-adapted approach to upfront G-CSF use rather than universal primary prophylaxis for all patients. In routine practice, decisions regarding primary prophylaxis should incorporate individual FN risk factors, including prior treatment burden, while patients not selected for upfront prophylaxis may be managed with close monitoring, timely secondary G-CSF, and standard-dose modifications as clinically indicated. Further prospective studies are warranted to refine patient selection and define the most effective timing and duration of growth-factor support.

Author contribution

MPü (Conceptualization, Data curation, Investigation, Methodology, Project administration, Supervision, Writing—original draft), AP-G (Investigation, Methodology, Writing—review & editing), AK (Investigation, Writing—review & editing), MPi (Investigation, Methodology, Writing—review & editing), JŻ (Investigation, Writing—review & editing), MH (Investigation, Project administration, Writing—review & editing), RS (Investigation, Writing—review & editing), HŠ (Investigation, Writing—review & editing), MM (Investigation, Writing—review & editing), AM (Investigation, Writing—review & editing), KW-S (Investigation, Writing—review & editing), ML-H (Investigation, Writing—review & editing), AP (Investigation, Writing—review & editing), DK (Investigation, Writing—review & editing), JŠ (Investigation, Writing—review & editing), IK (Investigation, Writing—review & editing), ID (Investigation, Writing—review & editing), MS-R (Investigation, Writing—review & editing), TC (Investigation, Writing—review & editing), LR (Investigation, Writing—review & editing), BC-A (Investigation, Writing—review & editing), AŁ (Investigation, Writing—review & editing), RP-M (Investigation, Writing—review & editing), MJ (Investigation, Writing—review & editing), ZB (Investigation, Project administration, Supervision, Writing—review & editing), and MK (Conceptualization, Formal analysis, Methodology, Supervision, Writing—review & editing). All authors reviewed and approved the final version of the manuscript.

Funding

The authors confirm that this study, including the preparation of the manuscript and the decision to submit it for publication, was carried out independently and received no financial support, grants, or external funding.

Data availability

The data generated and analyzed during the present study are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest

Mirosława Püsküllüoğlu received travel grants and lecture honoraria from AstraZeneca, Roche, Novartis, Eli Lilly, Janssen, Gilead, and Amgen. Anna Polakiewicz-Gilowska received lecture honoraria and travel support from AstraZeneca, Roche, Novartis, Eli Lilly, Swixx Biopharma, Gilead, and Pfizer. Aleksandra Konieczna received travel grants and lecture fees from Pfizer, Novartis, Eli Lilly, MSD, AstraZeneca, and Gilead; and advisory board honoraria from Novartis. Małgorzata Pieniążek received travel support and lecture fees from Pfizer and Novartis; lecture fees from Gilead; and participated in an advisory board for Novartis. Justyna Żubrowska received travel grants and lecture fees from MSD, Novartis, and AstraZeneca; and lecture fees from Pfizer, Swixx Biopharma, and Eli Lilly. Miloš Holánek received travel and conference support from MSD, Gilead, Eli Lilly, and AstraZeneca; received lecture and speaker’s fees from Novartis, Gilead, Eli Lilly, and Teva; and served on advisory boards for AstraZeneca, Eli Lilly, and Gilead. Renata Soumarová received honoraria and support from BMS, AstraZeneca, Pfizer, Novartis, Eli Lilly, MSD, Gilead, Janssen, Astellas, Servier, Accord, and GSK, including advisory board participation. Hana Študentová received speaker’s fees and conference support from Novartis, MSD, Janssen, Astellas, Pfizer, Richter Gedeon, AstraZeneca, and Ipsen. Miroslava Malejčíková served as a consultant and expert advisor for Novartis, Amgen, Pfizer, Roche, and AstraZeneca; received honoraria from Novartis, Pfizer, Roche, Teva, and AstraZeneca; and participated in clinical studies sponsored by Novartis, Roche, AstraZeneca, and Pfizer. Agnieszka Młodzińska received lecture honoraria from Novartis, Swixx Biopharma, and Gilead; received travel support from Eli Lilly, Pfizer, and Novartis. Karolina Winsko-Szczęsnowicz received travel support and lecture honoraria from Roche, Eli Lilly, Novartis, Pfizer, Gilead, MSD, and Aurovitas. Maja Lisik-Habib received honoraria for lectures from MSD, AstraZeneca, Johnson & Johnson, Roche, BMS, Astellas, Merck, and GSK. Anika Pękala declares no conflicts of interest. Daniel Krejčí received honoraria, congress fees, and travel support from MSD, AstraZeneca, Novartis, Eli Lilly, Roche, and Pfizer. Jan Šustr received conference support from Gilead and MSD; and speaker’s fees from AstraZeneca and MSD. Iveta Kolářová received honoraria and support from AstraZeneca, Novartis, MSD, Eli Lilly, Roche, Swixx Biopharma, and Gilead, including advisory board participation. Iwona Danielewicz declares no conflicts of interest. Magdalena Szymanik-Resko declares no conflicts of interest. Tomasz Ciszewski received travel grants and lecture fees from Pfizer, Novartis, Eli Lilly, MSD, Roche, Gilead, and AstraZeneca. Lenka Rušinová declares no conflicts of interest. Bogumiła Czartoryska-Arłukowicz received travel grants and lecture fees from Eli Lilly, Novartis, Pfizer, AstraZeneca, Gilead, and Roche. Aleksandra Łacko received honoraria and support from AstraZeneca, Pfizer, Novartis, Eli Lilly, Roche, and Gilead Sciences, including advisory board participation. Renata Pacholczak-Madej received travel grants from Accord and BMS; and lecture fees from AstraZeneca, BMS, MSD, GSK, Novartis, and Roche. Michał Jarząb received conference support from Gilead and Roche; speaker’s fees from Novartis, Roche, Eli Lilly, Pfizer, Teva, Exact Sciences, and Mammotome; and served on advisory boards for Novartis and Pfizer. Zuzana Bielčiková received travel grants and lecture honoraria from AstraZeneca, MSD, Roche, Novartis, Eli Lilly, and Gilead. Marcin Kubeczko served on advisory boards for Novartis; received speaker’s fees and conference support from Novartis, Roche, MSD, Pfizer, Eli Lilly, Teva, Amgen, Swixx Biopharma, AstraZeneca, and Gilead; and participated in clinical trials sponsored by Roche, MSD, Novartis, Seagen, and Gilead.