Study at a Glance
- The TELEGRAM study showed that, in the pooled analysis of the first three chemotherapy cycles, the incidence of musculoskeletal pain (MSP) was lower in the tuppefilgrastim group than in the group receiving other long-acting G-CSFs (14.4% vs. 28.0%, P=0.019).
- Among records in which MSP occurred, the tuppefilgrastim group had lower pain scores and a more favorable severity distribution, with no cases of severe MSP recorded during the observation period.
- By evaluating neutrophil dynamics alongside patients’ pain experiences, the study provides prospective data to inform pain management during long-acting G-CSF supportive treatment in breast cancer.
Introduction
Chemotherapy-induced neutropenia (CIN) and febrile neutropenia (FN) are important complications that need to be prevented during breast cancer chemotherapy. The use of long-acting granulocyte colony-stimulating factors (G-CSFs) helps prevent these complications and supports the delivery of chemotherapy. At the same time, reducing treatment-related pain has become an important consideration in optimizing supportive care.
At the 2026 European Society for Medical Oncology (ESMO) Congress, the TELEGRAM study conducted by Professor Wei Li’s team at Jiangsu Province Hospital was accepted for presentation (abstract number: 3950P; study number: NJMU-BC13; registration number: NCT07806084). The study explored pain outcomes associated with different long-acting G-CSFs and their relationship with changes in neutrophil counts. This article introduces the study design and key findings and discusses their clinical implications.
Study Background
Musculoskeletal pain is an adverse event of particular interest during long-acting G-CSF treatment. Its occurrence may involve several mechanisms, including bone marrow expansion, peripheral nociceptor sensitization, inflammation and immune regulation, and changes in bone metabolism. In patients with breast cancer, pain caused by taxane-containing chemotherapy may also overlap temporally with G-CSF-related pain, making pain identification and management more complex. Therefore, in addition to monitoring neutrophil recovery, systematically evaluating the incidence and severity of pain is an important component of comprehensive supportive care assessment.
Current management strategies for G-CSF-related pain include symptomatic interventions such as nonsteroidal anti-inflammatory drugs (NSAIDs), while optimizing administration timing has also attracted increasing attention. The prospective, randomized phase III PAINLESS study showed that, among patients with breast cancer receiving chemotherapy, administering pegfilgrastim 72 hours after chemotherapy, compared with administration at 24 or 48 hours, reduced the area under the curve of pain scores and the incidence of severe bone pain during the first cycle. No significant between-group difference in the incidence of neutropenia was observed. The NCCN Hematopoietic Growth Factors Guidelines, Version 1.2027, also cite this study, providing a reference for optimizing pain management through adjustments to treatment delivery.
In addition to administration timing, the drug exposure characteristics of long-acting G-CSFs may also affect pain burden. Studies have shown that G-CSF-related bone pain involves multiple mechanisms, including expansion of hematopoietic cells in the bone marrow, inflammatory cytokine release, and peripheral nerve sensitization. Previous dose-exploration studies have also observed a trend toward a higher incidence of bone pain with increasing G-CSF doses.
Tuppefilgrastim uses 40 kDa Y-shaped branched polyethylene glycol (PEG) modification. By increasing its hydrodynamic volume and delaying drug clearance, this formulation achieves relatively sustained drug activity at a lower administered dose. A phase II pharmacokinetic study showed that, compared with conventional 20 kDa PEG-rhG-CSF, tuppefilgrastim demonstrated a lower peak plasma concentration (Cmax: 14.95 vs. 122.27 ng/mL) and a relatively longer elimination half-life (3.34 vs. 2.73 days) at the respective study doses, exhibiting a “low peak, long-acting” pharmacokinetic profile. This characteristic provides a pharmacological hypothesis that reducing the high peak exposure associated with short-term drug exposure may help lessen bone marrow stimulation and pain burden. However, whether this translates into better clinical pain outcomes requires further validation.
Against this background, the TELEGRAM study (NJMU-BC13) prospectively compared the incidence and severity of MSP, as well as neutrophil recovery kinetics, between tuppefilgrastim and other long-acting G-CSFs in patients with breast cancer. The study aimed to explore the relationship between different treatment choices and patients’ pain experiences.
Study Design
TELEGRAM was a prospective, multicenter, nonrandomized controlled study enrolling patients with breast cancer who were receiving 21-day chemotherapy cycles containing anthracyclines and/or taxanes and who had an intermediate-to-high risk of FN.
Patients received prophylactic treatment with either tuppefilgrastim or another long-acting G-CSF selected by the investigator. Both groups received treatment 24–48 hours after chemotherapy. Tuppefilgrastim was administered once per cycle at a dose of 2 mg.
The study observed the first three chemotherapy cycles. Blood samples were collected and relevant assessments performed on days 0, 3, and 7 of each cycle. Pain severity was assessed using a 0–10 visual analogue scale (VAS), while absolute neutrophil count (ANC), adverse events, and concomitant medications were also recorded. The primary observations included MSP, changes in ANC, efficacy, and safety across the first three chemotherapy cycles.
The study planned to enroll 132 patients. The interim analysis presented at ESMO included 85 patients and 197 chemotherapy cycles, comprising 45 patients in the tuppefilgrastim group and 40 in the control group.
Study Results
No Significant Correlation Between ANC and Pain Scores, Highlighting the Complexity of Pain Mechanisms
The study showed statistically significant differences in ANC distributions across the first three chemotherapy cycles (overall P<0.05), with ANC levels in cycle 2 higher than those in cycle 1 (P<0.01). However, among the analysis records in which MSP occurred, no significant linear correlation was observed between ANC and VAS pain scores (r=0.03, P=0.87) (Figure 2).
Lower Incidence of MSP With Tuppefilgrastim, With Consistent Trends Across Three Consecutive Chemotherapy Cycles
The pooled analysis of the first three chemotherapy cycles, using the chemotherapy cycle as the unit of analysis, showed an MSP incidence of 14.4% in the tuppefilgrastim group, significantly lower than the 28.0% observed in the group receiving other long-acting G-CSFs (P=0.019), an absolute difference of 13.6 percentage points.
Further analysis by individual chemotherapy cycle showed that the MSP incidence in the tuppefilgrastim and control groups was 17.8% versus 27.5% in cycle 1, 15.6% versus 25.0% in cycle 2, and 7.4% versus 32.0% in cycle 3. The incidence was consistently lower in the tuppefilgrastim group across all three cycles (Figure 1).
Lower MSP Pain Scores and a Higher Proportion of Mild Pain With Tuppefilgrastim
Among the analysis records in which MSP occurred, the mean VAS score was 2.67±1.23 in the tuppefilgrastim group, significantly lower than the 4.65±2.31 observed in the control group (P<0.001). The median VAS scores were 2 and 4, respectively.
Pain was classified according to VAS score as mild (1–3), moderate (4–6), or severe (7–10). The proportion of mild MSP was 80.0% in the tuppefilgrastim group, higher than the 42.3% in the control group. The proportions of moderate MSP were 20.0% and 34.6%, respectively, while the proportions of severe MSP were 0% and 23.1%, respectively. The overall distribution of pain severity differed significantly between the groups (P=0.038). Notably, no cases of severe MSP were recorded in the tuppefilgrastim group among the MSP-positive analysis records in this interim analysis (Figure 2).
Comparable CIN Prevention, With No New Safety Signals
The study reported comparable CIN prevention outcomes between the two groups. Gastrointestinal adverse events were less frequent in the tuppefilgrastim group, and no new safety signals were identified.
Potential Relationship Between the Immune Microenvironment and ANC Recovery Kinetics Remains Under Investigation
Beyond pain outcomes, the TELEGRAM study also explored the relationship between the immune microenvironment and neutrophil recovery kinetics. Preliminary analyses suggested that changes in ANC may be associated with immune cytokines such as IL-4 and IL-17. Related investigations are ongoing. Further data are anticipated to provide additional insights into the processes underlying G-CSF-mediated neutrophil recovery.
Study Conclusions
The interim analysis of the TELEGRAM study (NJMU-BC13) showed that, compared with other long-acting G-CSFs, tuppefilgrastim was associated with a lower incidence of MSP. Among the analysis records in which MSP occurred, pain scores and the distribution of pain severity were also more favorable. Meanwhile, the overall CIN prevention outcomes were comparable between the two groups.
These findings provide new clinical reference data for pain management and individualized treatment selection during long-acting G-CSF supportive care in patients with breast cancer. Given the nonrandomized study design, these results require further validation as enrollment and follow-up continue.
Investigator Commentary
Preventing neutropenia and FN during breast cancer chemotherapy is essential to ensuring that treatment can be delivered as planned. With the widespread use of long-acting G-CSFs, we need to focus not only on their hematologic protective effects but also on patients’ treatment experiences. Musculoskeletal pain is an adverse reaction of particular concern during G-CSF treatment. It may affect patients’ daily lives and their willingness to continue subsequent supportive treatment. Therefore, reducing pain burden while effectively preventing CIN is an important issue in the refinement of supportive care for breast cancer.
One distinctive feature of the TELEGRAM study is that it prospectively evaluated the incidence and severity of pain alongside changes in ANC. In this interim analysis, the tuppefilgrastim group had a lower incidence of MSP in the pooled analysis of the first three chemotherapy cycles. Among the analysis records in which MSP occurred, pain scores and the distribution of pain severity were also more favorable. These findings suggest that, when evaluating long-acting G-CSF supportive treatment, pain outcomes can provide valuable information in addition to hematologic indicators, helping clinicians gain a more comprehensive understanding of patients’ actual treatment experiences.
Of particular interest, among patients who experienced MSP, no significant correlation was observed between ANC and VAS scores. This reminds us that G-CSF-related pain may involve multiple mechanisms, including bone marrow hematopoietic activity, inflammatory responses, and neural sensitization. Pain severity therefore cannot be judged simply by peripheral blood neutrophil counts. The distinctive PEG structure and pharmacokinetic characteristics of tuppefilgrastim also provide a basis for further investigating the relationship between drug exposure profiles and pain outcomes, although the specific mechanisms remain to be validated.
Meanwhile, the previous PAINLESS study and relevant guideline updates suggest that administration timing is also an important factor in managing G-CSF-related pain. TELEGRAM used a nonrandomized design, and differences existed between the two groups in patient characteristics and the actual distribution of administration times. Therefore, the observed differences in pain outcomes should be interpreted cautiously, taking these factors into account. As enrollment, follow-up, and further analyses progress, we look forward to a more comprehensive assessment of the benefits and treatment burdens associated with different long-acting G-CSFs. This may provide further evidence to support individualized supportive care that balances effective CIN prevention with patients’ treatment experiences.
Professor Wei Li
