Dual-target neoadjuvant therapy with denosumab and sunitinib for giant cell tumor of bone: a sarcoma center experience

Abstract

Background

Giant cell tumor of bone (GCTB) is a locally aggressive tumor characterized by osteoclast-like giant cells and neoplastic stromal cells. While denosumab effectively targets giant cells, stromal cells may persist and drive recurrence. This study evaluates the efficacy and safety of a novel neoadjuvant regimen combining denosumab with sunitinib, a tyrosine kinase inhibitor targeting stromal cell signaling. To our knowledge, this represents the first clinical series evaluating a novel neoadjuvant regimen combining denosumab with sunitinib in GCTB.

Methods

We conducted a retrospective case series of 16 patients with histologically confirmed Campanacci grade 3 (or grade 2 in anatomically challenging locations) GCTB treated at a tertiary sarcoma center between January 2019 and September 2024. All patients received neoadjuvant denosumab and sunitinib prior to definitive surgery. Surgical outcomes, local recurrence, histopathological tumor regression, disease-free survival, and the relative tyrosine phosphorylation levels were analyzed.

Results

Of the 16 patients (mean age 28.7 years), most had Campanacci grade 3 lesions (87.5%). En-bloc resection was performed in 43.8% and curettage in 50% of cases. Histological analysis showed near-complete regression of giant cells (mean 99.7%) and partial regression of stromal cells (mean 62.8%). Two patients (12.5%) experienced local recurrence. Disease-free survival at two years was 80%. Treatment was well tolerated, with no unexpected adverse events or dose-limiting toxicities.

Conclusion

Neoadjuvant combination therapy with denosumab and sunitinib was feasible and well tolerated and resulted in substantial histological regression of both giant cells and neoplastic stromal cells. While recurrence outcomes in this limited cohort remain descriptive, the observed stromal suppression and RTK signaling profiles support the biological rationale for dual-targeted therapy. These findings justify further prospective investigation, particularly in aggressive or surgically challenging GCTB cases.

1. Introduction

According to the 2020 World Health Organization (WHO) classification, giant-cell tumor of bone (GCTB) is classified as an intermediate malignant primary bone tumor. This classification reflects its potential for aggressive behavior and a low tendency to metastasize, mainly to the lungs [1], [2]. GCTB is a relatively common diagnosis, which accounts for approximately 4–10% of all primary bone tumors [3], [4], [5]. The peak incidence is between the ages of 20 and 40 years. A solitary occurrence of GCTB is typical, with a predilection for the meta-epiphyseal region of long bones (85%)[6], [7], [8]. Predominant localization is around the knee, as it affects mostly distal femur and proximal tibia, with subsequent involvement of the distal radius [6], [8], [9], [10]. Recurrence rates vary widely and are usually reported between 12% and 65%[5], [11]. The numbers strongly depend on the type of selected surgical procedure and the neoadjuvant use of denosumab [12], [13], [14]. GCTB is histologically composed of osteoclast-like giant cells and neoplastic mononuclear stromal cells. Stromal cells express receptor activator of nuclear factor kappa B ligand (RANKL) which stimulates osteoclast activation through the RANKL–RANK signaling pathway. As a result, osteoclast-like cells are secondarily recruited into the tumor site and are responsible for its aggressive osteolytic activity [15], [16].

According to the Campanacci classification, the course of treatment should be selected. Surgical treatment options include mainly intralesional curettage or en-bloc resection and the choice is made based on the extent of the local disease as well as tumor’s location and patient́s overall health and preferences [17]. Curettage is associated with a higher risk of local recurrence when performed alone [10], [18], [19]. Since it has been reported to reduce local recurrence rate, it is recommended to do aggressive curettage with a high-speed burr as well as to include the use of local adjuvants during the procedure. Among others, for example phenol, liquid nitrogen, ethanol or electrocautery [12], [20], [21], [22].To reduce the risk of local recurrence, systemic treatment with bisphosphonates, interferon alpha (IFN-α) or denosumab is also being used [23], [24], [25]. For grade 3 cases of GCTB according to Campanacci classification with soft tissue involvement and bone destruction and for recurrent disease, en-bloc resection should be considered [17], [18], [26], [27]. Resection with wide margins reduces a chance of tumor recurrence however is associated with worse functional results as well as possible complications in reconstruction [10].

Denosumab is a fully human monoclonal antibody against RANKL that leads to elimination of osteoclast-like giant cells because of interrupting RANK-RANKL binding which is necessary for osteoclast formation. While denosumab is highly effective in eliminating osteoclast-like giant cells and facilitating local disease control, denosumab-assisted surgery, particularly curettage, remains associated with an increased risk of local recurrence in selected clinical scenarios. This has been attributed to the persistence of viable neoplastic stromal cells, which represent a known biological limitation of denosumab [25], [26], [27]. Sunitinib, a multi-targeted RTK inhibitor, has been investigated as a potential adjunct to denosumab in the management of GCTB [28]. Among other receptors, it targets especially platelet-derived growth factor receptors (PDGFRs) and vascular endothelial growth factor receptors (VEGFRs) leading to inhibition of neoplastic stromal cells. Therefore, denosumab treatment complemented with sunitinib targets osteoclast-like giant cells as well as neoplastic stromal cells at the same time [28], [29].

The aim of this study was to assess the efficacy and safety of neoadjuvant treatment with a combination of denosumab and sunitinib in patients with GCTB. Specifically, the study sought to evaluate the impact of this therapeutic approach on surgical management, local recurrence rates, and disease-free survival. In addition, the study aimed to investigate the histopathological response of the tumor—particularly the behavior of multinucleated giant cells and neoplastic stromal cells—following combination therapy as well as the relative tyrosine phosphorylation levels of 49 receptor tyrosine kinases (RTKs). The rationale for exploring additional stromal targeting is therefore not to escalate therapy routinely, but to address specific high-risk or surgically challenging cases in which recurrence remains a concern. In this context, sunitinib was explored as a biologically informed adjunct aimed at modulating stromal cell activity, rather than as a replacement for or routine addition to standard therapy. To our knowledge, this represents the first clinical series evaluating a novel neoadjuvant regimen combining denosumab with sunitinib in GCTB.

2. Methods

2.1. Study Design and Patient Selection

This was a retrospective cohort study conducted at the First Department of Orthopaedic Surgery, St. Anne’s University Hospital in Brno, Czechia. A total of 16 patients diagnosed with GCTB were treated between January 2019 and September 2024. The diagnosis of GCTB was confirmed histologically using hematoxylin-eosin staining and immunohistochemistry and independently reviewed by a dedicated musculoskeletal pathologist. The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Research Ethics Committee of Masaryk University (Brno, Czechia) (Protocol Code: EKV2021074)

Baseline characteristics of the cohort are summarized in Table 1. The mean age at inclusion was 28.7 ±13.5 years, with a predominance of female patients (75%). Of the 15 patients with available data, 14 had no comorbidities, and none presented with metastases or misdiagnoses at the time of initial evaluation. The mean duration of symptoms before diagnosis was 3.4±2.7 months (Table 1).

Table 1.

Characteristics of patients with GCTB treated with the neoadjuvant treatment consisted of denosumab in combination with sunitinib.

Feature	Value
Overall (n)	16
Age at inclusion, years (SD)	28.7 (±13.5)
Sex, n (%) Male Female	4 (25%) 12 (75%)
Follow-up, months (SD)	31.3 (±17.8)
Symptom duration, months (SD)	3.4 (±2.7)
Previous surgery, n (%) Yes No	2 (12.5%) 14 (87.5%)
Metastasis at diagnosis, n (%) Yes No	0 16 (100%)
Previous misdiagnosis, n (%) Yes No	0 16 (100%)

Tumor grading revealed that the majority had Campanacci grade 3 lesions (87.5%), indicating aggressive radiographic features. Lesion locations were variable, with the radius (31.3%), ulna (18.7%), and tibia (12.5%) being the most commonly affected bones. Other sites included the hand, foot, fibula, humerus, and sacrum. The mean follow-up time was 31.3 (±17.8) months, with a minimum of one year (Table 2).

Table 2.

Grading, size and site of GCTB.

Feature	Value
Campanacci grade, n (%) G1 G2 G3	0 2 (12.5%) 14 (87.5%)
Tumor size, cm (SD)	6.7 (±2.6)
Tumor site, n (%) Hand Foot Radius Ulna Tibia Fibula Humerus Sacrum	2 (12.5%) 1 (6.3%) 5 (31.3%) 3 (18.7%) 2 (12.5%) 1 (6.3%) 1 (6.3%) 1 (6.3%)

Neoadjuvant therapy was recommended following a comprehensive review by the institutional musculoskeletal oncology committee, which assessed each case through a multidisciplinary evaluation of clinical, radiological, and histopathological findings. Patients with Campanacci grade 3 GCTB, or grade 2 with anticipated surgically difficult location, treated with neoadjuvant denosumab and sunitinib between 2019–2024, were included. Patients were selected taking into account the presence of extraosseous extension, degree of osteolysis, and aggressiveness of the stromal cell population, which was assessed in vitro. All cases were treated in accordance with institutional ethical guidelines and with informed consent.

2.2. Treatment Protocol

The neoadjuvant treatment protocol consisted of denosumab in combination with sunitinib, initiated before definitive surgery. The rationale for the addition of sunitinib was based on prior in vitro analysis indicating heightened activation of RTKs in the neoplastic stromal cells, supporting a synergistic antitumor effect.

Prior to therapy initiation, all patients underwent a pre-treatment evaluation, including dental examination to exclude infectious foci, and echocardiography to assess baseline cardiac function.

Denosumab was administered at a dose of 120 mg subcutaneously every four weeks, with two additional loading doses on days 8 and 15 of the first month. Patients were supplemented with calcium and cholecalciferol throughout the treatment period to prevent hypocalcemia.

Sunitinib was introduced during the final month (30 days) of neoadjuvant therapy at a daily oral dose of 25 mg. This staggered approach limited overall exposure to sunitinib, allowed assessment of tolerability in close temporal proximity to surgery, and minimized potential perioperative risks associated with tyrosine kinase inhibition, including effects on wound healing and vascular integrity. In addition, sunitinib was administered after initial denosumab therapy because denosumab induces biological and immunophenotypic changes in neoplastic stromal cells [28], which may increase their susceptibility to subsequent RTK inhibition while permitting a limited and safer preoperative exposure.

During the course of treatment, patients were closely monitored with weekly blood pressure measurements, serial electrocardiograms to detect arrhythmias or QT interval prolongation, liver function tests to screen for hepatotoxicity, and serum calcium levels to detect hypocalcemia. Adverse events were assessed retrospectively and graded according to the Common Terminology Criteria for Adverse Events (CTCAE), version 5.0.

All surgical procedures were performed at a tertiary sarcoma center by experienced orthopedic oncologic surgeons following standardized institutional protocols. Intralesional curettage included the use of a high-speed burr and local adjuvant measures where appropriate to minimize residual tumor burden [30]. Surgical intervention was planned based on clinical and radiographic response to therapy. The types of surgery performed included en-bloc resection, intralesional curettage with PMMA augmentation, and in one case, no surgery due to individual clinical considerations.

2.3. Clinical and Radiological Follow-Up

Postoperative follow-up was standardized and included clinical and radiological evaluations. Patients were assessed every three months for the first two years, every six months up to five years, and annually thereafter. Follow-up imaging included plain radiographs and MRI or CT scans.

3. Evaluation

3.1. Oncological Outcomes

Treatment efficacy was assessed through both radiological and clinical surveillance to determine local tumor control and disease-free survival. Local recurrence was defined as radiological and/or histological evidence of tumor regrowth at the original site following surgical treatment. Time to recurrence was calculated from the date of surgery to the date of radiologically confirmed recurrence. Disease-free survival was measured from the date of surgery to either the date of recurrence or the last follow-up for patients without recurrence. Survival analysis was performed using Kaplan–Meier methodology. Disease-free survival was only calculated for those who had surgery. One patient did not proceed to surgery; for this case, response was assessed radiologically and they were excluded from calculations.

3.2. Histopathological Response Assessment

The degree of tumor regression was assessed in post-resection specimens by an experienced musculoskeletal pathologist. Specimens were evaluated using hematoxylin-eosin staining and immunohistochemistry. Regression percentages for giant and stromal cells were determined by visual estimation and image-assisted quantification of viable cellular areas on hematoxylin-eosin–stained sections, expressed as the proportion of residual tumor cells relative to the pre-treatment baseline. Stromal cell regression was defined specifically as a reduction in the proportion of viable H3G34W-positive neoplastic stromal cells. This parameter reflects a composite histological response, treatment-induced differentiation, and/or loss of neoplastic cellularity, rather than a direct measure of cytotoxicity or apoptosis. Histological response was therefore used as a surrogate marker of biological treatment effect rather than a definitive indicator of clinical efficacy.

Complete radiological or histological response was not defined as an endpoint in this study. Given the characteristic therapy-induced changes in GCTB, including osteosclerosis, fibrous tissue proliferation, and reactive osteoplasia, standard definitions of complete response were not applicable. Treatment response was therefore assessed based on the degree of histological regression of H3G34W-positive neoplastic stromal cells relative to pre-treatment biopsy specimens.

3.3. Phospho-RTK arrays

Phospho-RTK analysis was performed on fresh-frozen tumor tissue samples obtained from all 16 patients included in the study cohort. Tissue samples were collected intraoperatively. The analyzed samples therefore fully reflect the clinicopathological characteristics of the study population. The relative tyrosine phosphorylation levels of 49 RTKs were analyzed in fresh-frozen tumor tissue samples using the Proteome Profiler™ Human Phospho-RTK Array Kit (cat. no. ARY001B; R&D Systems, Minneapolis, MN, USA) as previously described [31]. Chemiluminescent signals were quantified using ImageJ software [32]. Data are presented as a heat map. Sixteen RTKs with the highest mean of pixel density were selected for the analysis. Hierarchical clustering was applied to the kinase phosphorylation data to identify potential characteristic patterns in kinase phosphorylation status after combination therapy. All detected proteins are listed in the Supplementary Materials (Table S1).

3.4. Patient Cohort Summary and Representative Clinical Cases

A summary of all cases is presented in the Table 4 in the Results section. To illustrate clinical decision-making, treatment complexity, and patient outcomes, three representative cases were selected from the study cohort. These cases reflect diversity in anatomical site, extent of disease, surgical approach, and response to the neoadjuvant therapy. Each case includes a brief description of the rationale for combination therapy, surgical management, histologic findings, and functional recovery.

Table 4.

Summary of individual details of each patient.

Patient No.	Age (years)	Tumor site	Campanacci grade	Type of surgery	Follow-up (months)	Local recurrence	Time to LR (months)	Giant-cell regression (%)	Stromal cell regression (%)	CTCAE severity scale
1	18	Foot	3	Curettage	69	No		100	100
2	19	Radius	3	En-bloc resection	57	No		100	100
3	20	Ulna	3	En-bloc resection	41	No		100	90	1
4	23	Sacrum	3	N/A	57	No		100	99	1
5	15	Hand	3	En-bloc resection	24	No		100	70
6	50	Hand	3	Curettage	37	No		100	85
7	45	Tibia	3	Curettage	42	No		100	10
8	18	Ulna	3	Curettage	36	No		100	70
9	16	Ulna	3	Curettage	24	Yes	24	100	10
10	52	Tibia	2	Curettage	9	Yes	9	95	60
11	36	Radius	3	Curettage	21	No		95	95
12	25	Radius	3	En-bloc resection	18	No		100	40
13	26	Fibula	3	En-bloc resection	18	No		100	30
14	23	Humerus	3	En-bloc resection	18	No		100	40
15	21	Radius	2	Curettage	15	No		100	30
16	52	Radius	3	En-bloc resection	15	No		100	70	1

3.5. Statistical Analysis

Descriptive statistics were used to summarize baseline characteristics, treatment outcomes, and follow-up data. Continuous variables were reported as mean ± standard deviation (SD) and categorical variables as frequencies and percentages. Disease-free survival was estimated using Kaplan–Meier analysis. Statistical analyses were performed using R software (version 4.4.2) (R Foundation for Statistical Computing, Vienna, Austria) within the RStudio development environment.

4. Results

4.1. Surgical Management and Recurrence

The combination was generally well tolerated; no patients required dose reductions or delays. Treatment-related adverse events were mild and graded as CTCAE 1 and occurred in 3 cases. Out of 16 patients treated with neoadjuvant denosumab and sunitinib, the majority underwent curettage with polymethyl methacrylate (PMMA) augmentation (8/16; 50%), followed by en-bloc resection (7/16; 43.8%). The one patient with sacrum infiltration (6.3%) who did not undergo surgery (due to the size of tumor and the potential postoperative morbidity) remained on prolonged denosumab therapy, with stable disease and no evidence of metastasis at last follow-up (57 months). Local recurrence occurred in two patients (12.5%) at 9 and 24 months following surgery. (Table 3).

Table 3.

Surgical management and local recurrence of GCTB treated with the neoadjuvant treatment consisted of denosumab in combination with sunitinib.

Feature	Value
Type of surgery n (%) En-bloc resection Curettage + PMMA No surgical therapy	7 (43.8%) 8 (50%) 1 (6.3%)
Local recurrence, n (%) Yes No	2 (12.5%) 14 (87.5%)
Time to local recurrence, months Patient No. 9 Patient No. 10	24 9

4.2. Disease-Free Survival

The majority of patients maintained durable tumor control following treatment. At one year after surgery, 93.3% (95% CI 63.2–99.1) of patients remained free of local recurrence, with 13 patients still at risk. By the second year, 80% (95% CI 42.3–95.5) of patients continued to be disease-free, based on 7 patients at risk during that period. This disease-free rate remained stable into the third year, with 80% (95% CI 42.3–95.5) of patients showing no signs of recurrence, now based on 4 patients at risk. Disease-free survival was illustrated using Kaplan–Meier to provide a visual overview of follow-up duration and timing of recurrence events (Fig.1).

Fig. 1.

Disease-Free Survival of patients with GCTB treated with the neoadjuvant treatment consisted of denosumab in combination with sunitinib.

4.3. Histopathological Tumor Regression

Postoperative histological evaluation showed near-complete elimination of multinucleated giant cells, with an average regression of 99.7% (±1.3). Stromal cell regression was more variable, averaging 62.8% (±32.2), indicating partial sensitivity of neoplastic stromal cells to the neoadjuvant regimen and highlighting biological heterogeneity across tumors. Notably, the two tumors that developed local recurrence exhibited stromal regression of only 60% and 10%, respectively, compared with the overall mean of 62.8%, suggesting a potential association between insufficient stromal suppression and recurrence risk.

4.4. RTK phosphorylation profiling

To analyze the phosphorylation profiles of RTKs in tumors after combination therapy with denosumab and sunitinib, phosphoprotein arrays were employed. Our data obtained from fresh-frozen tumor samples showed that, in most GCTB tissues, the highest relative phosphorylation was observed for Epidermal growth factor receptor (EGFR), followed by lower and variable phosphorylation of Platelet-derived growth factor receptor beta (PDGFRβ), both insulin receptors (InsR and IGF-IR), Platelet-derived growth factor receptor alpha (PDGFRα), and Macrophage colony-stimulating factor receptor (M-CSFR). The occurrence of increased phosphorylation of other receptors was sporadic (Fig. 2).

Fig. 2.

Phosphorylation analysis of RTKs in patient-derived fresh-frozen GCTB samples.

5. Patient Cohort Summary and Representative Clinical Cases

Selected representative cases are presented below to illustrate treatment response across different anatomical sites and surgical approaches (see Table 4 for full cohort summary).

Patient No. 11 (Table 4): A 36-year-old female with Campanacci grade 3 GCTB of the distal radius was treated with neoadjuvant denosumab followed by sunitinib prior to intralesional curettage with PMMA augmentation (Fig.3). Histopathological evaluation demonstrated approximately 95% regression of both multinucleated giant cells and neoplastic stromal cells (Fig.4). At 21 months of follow-up, the patient remained disease-free with good functional outcome (Fig.5).

Fig. 3.

Pre-therapy: GCTB tissue with abundant mononuclear stromal tumor cells (green arrows) and osteoblast-like elements (white arrow); tumor dominates the biopsy sample (left). Extensive H3G34W nuclear positivity in stromal cells; a few H3G34W-negative multinucleated cells are also present (right).

Fig. 4.

Post-therapy: Tumor tissue is largely replaced by reactive bone (left). H3G34W staining shows complete loss of giant cells and a marked reduction of H3.3-positive tumor cells, with reactive osteoplsia present throughout the sample (right).

Fig. 5.

Representative radiological course of distal radius GCTB. (A) Pre-treatment Campanacci grade 3 giant cell tumor of bone of the distal radius. (B) Post-neoadjuvant imaging after denosumab and sunitinib showing sclerosis and structural remodeling. (C) Postoperative imaging after intralesional resection and PMMA augmentation at 21 months of follow-up.

Patient No. 13 (Table 4): A 26-year-old female with Campanacci grade 3 GCTB of the fibula (Fig.6) underwent neoadjuvant therapy with denosumab and sunitinib followed by en-bloc resection without reconstruction. Histological analysis showed complete regression of giant cells (100%) and partial regression of stromal cells (30%) (Fig.7). At 18 months of follow-up, no local recurrence was observed and functional outcome was favorable.

Fig. 6.

Pre-therapy: Numerous tumor cells with scattered multinucleated giant cells – white arrow (left). H3G34W staining in the same case demonstrates nuclear positivity in tumor cells (right).

Fig. 7.

Post-therapy: The GCTB specimen shows reactive osteoplasia, complete absence of giant cells, and partial depletion of stromal cells (left). H3G34W staining confirms the continued presence of stromal tumor cells despite overall reduction (right).

Patient No. 7 (Table 4): A 45-year-old male with Campanacci grade 3 GCTB of the proximal tibia (Fig.8) received neoadjuvant denosumab and sunitinib followed by intralesional curettage with PMMA augmentation. Histopathological examination revealed complete regression of giant cells (100%) with limited stromal cell regression (10%) (Fig.9). At 42 months of follow-up, the patient remained free of local recurrence with preserved limb function.

Fig. 8.

Pre-therapy: GCTB sample showing abundant multinucleated (osteoclast-like) giant cells and mononuclear stromal cells (left). Tumor cells highlighted with H3G34W immunostaining; giant cells are negative (they stain positively with CD68, not shown) (right).

Fig. 9.

Post-therapy: Giant cells are entirely depleted (Denosumab), but mononuclear stromal cells persist. These resemble reactive fibroblasts, requiring immunohistochemistry (left). H3G34W stain reveals numerous positive stromal cells among negative fibroblasts. Decalcification (due to tumor calcification elsewhere, not shown) slightly reduced nuclear staining, but positivity remains diagnostic (right).

6. Discussion

This study evaluated a new potential neoadjuvant regimen combining denosumab and sunitinib for the treatment of GCTB with the aim of improving histological response and reducing recurrence by targeting both giant cells and stromal tumor cells. Our findings suggest that this dual-targeted approach is feasible and well-tolerated and may offer biological advantages over denosumab monotherapy. Given the limited cohort size, this study was not powered to statistically evaluate differences in local recurrence rates. Therefore, recurrence outcomes should be interpreted descriptively rather than as evidence of clinical superiority. Instead, the primary evidence supporting the combination approach arises from objective histopathological findings demonstrating partial but consistent suppression of neoplastic stromal cells, a tumor compartment not affected by denosumab alone.

Although GCTB most commonly occurs around the knee, forearm lesions were overrepresented in our cohort. Tumors in the distal radius and ulna are often diagnosed later due to initially subtle symptoms and delayed functional impairment and therefore tend to be more locally advanced at presentation. In addition, these anatomical locations are associated with less favorable surgical conditions and a higher risk of functional compromise, leading to more frequent indication for neoadjuvant treatment. Surgical technique and margin adequacy remain critical determinants of local control in GCTB and represent an important potential confounder in the interpretation of oncological outcomes. In our cohort, both recurrences occurred following intralesional curettage, consistent with previously reported patterns, and underscore the continued importance of surgical approach. The observed outcomes should be interpreted in the context of combined modality treatment rather than as an isolated effect of the neoadjuvant regimen. In this setting, the reduction of neoplastic stromal cells supports a biologically relevant treatment effect that is unlikely to be fully explained by surgical factors alone.

Denosumab has become a cornerstone in the neoadjuvant management of GCTB for its ability to eliminate osteoclast-like giant cells and is used in cases when radical resection cannot be performed. Administration before an en-bloc resection leads to osteosclerosis and lower probability of tumor spillage and therefore reduces local recurrence rate [10], [33]. In contrast, neoadjuvant denosumab prior to curettage has been associated with an increased risk of local recurrence compared with curettage alone [26], [27], [28], a finding that has been mechanistically linked to persistence of viable neoplastic stromal cells, which are not targeted by denosumab and are believed to drive recurrence [25], [26], [27]. In our series, the addition of sunitinib was associated with partial cytoreduction of neoplastic stromal cells, with a mean regression of 62.8%. While this stromal response was variable, likely reflecting tumor heterogeneity and differential RTK expression, it represents a biological effect not achievable with denosumab alone and provides hypothesis-generating support for a potential synergistic benefit.

In our cohort, phosphorylation analysis indicated a predominant activation of EGFR in many GCTB samples, whereas activation of other RTKs, including PDGFRs, insulin receptors, and M-CSFR, was generally weaker and more variable. Only occasional activation of additional receptors was detected, suggesting that RTK signaling in GCTB is driven by a limited set of pathways. This pattern points to EGFR, alongside PDGFR and VEGFR targeted by sunitinib, as a relevant therapeutic target in a subset of patients. Although EGFR phosphorylation was frequently observed, sunitinib was selected because multikinase inhibition has demonstrated greater functional suppression of stromal cell viability than single-target EGFR inhibition, particularly in the context of denosumab-altered RTK signaling [28], [34]. Taken together, these findings are exploratory and do not establish a direct mechanistic link to treatment response; however, they provide biologically relevant context for the observed stromal cell behavior. In particular, PDGFR and VEGFR signaling pathways are known to play key roles in stromal cell proliferation and tumor vascularization, and their inhibition by sunitinib may contribute to the partial suppression of neoplastic stromal cells observed in our cohort. Nevertheless, given the descriptive nature of this analysis, these results require further validation in functional and prospective studies.

Histologically, the nearly complete eradication of giant cells (mean 99.7%) is consistent with prior studies on denosumab [23]. The observed stromal cell regression following combination therapy, although partial, represents a promising step toward a more comprehensive therapeutic effect. This dual activity aligns with the known pharmacological targets of sunitinib, which inhibits PDGFR and VEGFR signaling pathways involved in stromal cell proliferation and tumor vascularization [28], [29]. The stromal regression observed in our cohort may reflect a reduction in the viable stromal cell pool or suppression of stromal cell activity through RTK inhibition. While this does not eliminate stromal cell dormancy, it may plausibly limit the reactivation potential of residual tumor cells, providing a biological explanation for the association between low stromal regression and recurrence. Although disease-free survival provides a clinically relevant outcome measure, it does not directly reflect treatment responsiveness and is influenced by surgical and other factors. Therefore, histopathological assessment, particularly the evaluation of neoplastic stromal cell regression, was considered a more appropriate indicator of biological treatment effect, as standard size-based criteria such as RECIST are not applicable in GCTB treated with denosumab and do not adequately capture treatment-induced biological changes.

The dual therapy and more aggressive surgical approaches when appropriate (e.g., higher rate of en-bloc resections) may have contributed to the observed disease-free survival rate of 80% at two years. Clinical outcomes following denosumab-based treatment of GCTB vary substantially across published studies, particularly in relation to the type of surgical management. Several reports have demonstrated that the use of denosumab in conjunction with intralesional curettage may be associated with an increased risk of local recurrence, with rates commonly reported in the range of approximately 20–50% depending on surgical technique, tumor location, and patient selection [13], [14]. In contrast, en-bloc resection is generally associated with lower recurrence rates, often below 20%, albeit at the cost of greater functional morbidity [18]. Larger multicenter analyses of denosumab-treated GCTB further underscore the heterogeneity of outcomes, reflecting variability in tumor biology, anatomical site, and treatment strategy [26]. In this context, the recurrence rate observed in our cohort (12.5%) appears comparable to or potentially lower than that reported in many existing series, despite the predominance of Campanacci grade 3 and anatomically challenging tumors in our population. Importantly, both recurrences in our series occurred following curettage, which is consistent with previously reported patterns and reinforces the continued importance of surgical approach in local disease control. Nevertheless, these comparisons remain indirect and should be interpreted with caution given the limited cohort size, lack of a control group, and heterogeneity across studies.

When compared with our previous institutional series of 95 surgically treated GCTB patients [10], in which the five-year recurrence-free survival was 85.6% after resection and 72.2% after curettage without neoadjuvant therapy, and only 57.1% in patients receiving denosumab plus curettage, the present cohort receiving dual therapy with denosumab and sunitinib demonstrated a lower rate of local recurrence and improved recurrence-free survival despite being limited to Campanacci grade 3 or otherwise surgically challenging cases. However, this comparison is descriptive and contextual in nature and does not constitute a controlled assessment of treatment efficacy. Rather, the observed differences provide preliminary support for the biological plausibility that additional stromal targeting may mitigate the increased recurrence risk.

From a safety perspective, the combination of denosumab and sunitinib was generally well tolerated. One patient developed a low-grade palmar–plantar skin reaction. Two patients experienced mild gastrointestinal symptoms, including nausea in one case and nausea with diarrhea in another. No grade ≥2 adverse events were observed, and no dose modifications, treatment interruptions, or specific supportive interventions were required. This is an important consideration, as sunitinib can be associated with hematologic, hepatic, and cardiovascular toxicities when used in other oncologic indications. The absence of significant complications in our cohort supports the regimen’s viability in a neoadjuvant setting, particularly when treatment duration is limited. Administration restricted to the neoadjuvant window may have minimized toxicity risk while still conferring therapeutic benefit.

This study has several limitations. Given the rarity of GCTB and the ethical as well as logistical challenges of assigning patients to placebo or withholding established treatment, a randomized controlled trial to evaluate this novel dual-targeted approach was not feasible. Consequently, clinical outcomes—particularly local recurrence—should be interpreted as descriptive and preliminary, and causal attribution of benefit to sunitinib cannot be established. Follow-up duration, while sufficient for early recurrence assessment, remains limited for long-term oncological outcomes. Additional biological analyses, such as assessment of stromal cell proliferation (Ki-67), apoptosis (cleaved caspase-3), or metabolic activity by PET-CT, could further clarify the nature of stromal response but were not performed due to limited archival tissue availability and the non-standardized use of PET imaging in GCTB during the study period. Standard response criteria such as RECIST and tumor regression grading systems were not applied, as denosumab-based therapy commonly induces osteosclerosis and morphological remodeling without proportional tumor size reduction and these systems have not been validated for GCTB. In this context, histopathological evaluation remains the most informative modality for treatment response assessment.

7. Conclusion

Neoadjuvant combination therapy with denosumab and sunitinib was feasible and well tolerated and resulted in substantial histological regression of both giant cells and neoplastic stromal cells. While recurrence outcomes in this limited cohort remain descriptive, the observed stromal suppression and RTK signaling profiles support the biological rationale for dual-targeted therapy. These findings justify further prospective investigation, particularly in aggressive or surgically challenging GCTB cases.

CRediT authorship contribution statement

Michal Mahdal: Writing – review & editing, Writing – original draft, Validation, Supervision, Resources, Methodology, Formal analysis, Conceptualization. Vasileios Apostolopoulos: Writing – review & editing, Writing – original draft, Methodology, Formal analysis, Conceptualization. Peter Múdrý: Writing – original draft, Visualization, Methodology, Data curation. Jakub Neradil: Writing – original draft, Visualization, Supervision, Methodology, Data curation. Marian Kubíček: Writing – review & editing, Writing – original draft, Visualization, Methodology. Dušan Zoufalý: Writing – review & editing, Writing – original draft, Visualization. Dagmar Adámková: Writing – review & editing, Supervision, Methodology, Data curation. Tomáš Tomáš: Writing – review & editing, Supervision, Methodology, Formal analysis, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: