Angiogenic and Stromal Remodeling Associated With Aggressive Fibrosarcomatous Dermatofibrosarcoma Protuberans

Sakura Shiraishi; Takeshi Iwasaki; Ken Takigawa; Kengo Kawaguchi; Hiroki Sonoda; Kohta Miyawaki; Takumi Tomonaga; Taro Mori; Makoto Endo; Koji Kato; Yasuharu Nakashima; Yoshinao Oda

doi:10.1111/cas.70558

. 2026 Oct 11:10.1111/cas.70558. Online ahead of print. doi: 10.1111/cas.70558

Angiogenic and Stromal Remodeling Associated With Aggressive Fibrosarcomatous Dermatofibrosarcoma Protuberans

Sakura Shiraishi ^1,², Takeshi Iwasaki ¹, Ken Takigawa ³, Kengo Kawaguchi ², Hiroki Sonoda ², Kohta Miyawaki ³, Takumi Tomonaga ¹, Taro Mori ¹, Makoto Endo ², Koji Kato ³, Yasuharu Nakashima ², Yoshinao Oda ^1,^✉

PMCID: PMC13655322 PMID: 42860657

ABSTRACT

Although collagen type 1 alpha 1 (COL1A1)–platelet‐derived growth factor B‐chain (PDGFB) gene fusion is a characteristic driver alteration in fibrosarcomatous dermatofibrosarcoma protuberans (FS‐DFSP), an aggressive DFSP variant, the tumor microenvironment (TME) underlying the aggressive phenotype remains unclear. We investigated microenvironmental and molecular differences between pure‐DFSP and FS‐DFSP using integrated histopathologic and transcriptomic analyses of archival samples from 122 patients, including 39 with FS‐DFSP. Histopathologic evaluation, immunohistochemistry, and transcriptomic analyses were performed to assess vascular features, cancer‐associated fibroblasts (CAFs), macrophage infiltration, and angiogenesis‐related molecular alterations. FS‐DFSP demonstrated increased CD31‐positive vascular structures without a corresponding increase in αSMA‐positive vascular structures, suggesting immature vascular features. CAF markers, including periostin and fibroblast activation protein, were significantly upregulated in intratumoral and invasive front regions of FS‐DFSP. Gene expression analysis was performed separately for the fibrosarcomatous and DFSP components of the FS‐DFSP cases and for the pure‐DFSP cases. By targeted transcriptomic profiling, the DFSP and fibrosarcomatous components within the FS‐DFSP cases shared molecular features, suggesting early tumor microenvironment‐associated alterations that preceded overt fibrosarcomatous morphology. Integrated transcriptomic and protein analyses revealed upregulation of angiogenesis‐related molecules, including vascular endothelial growth factor A (VEGFA), C‐C motif chemokine ligand 2 (CCL2), and C‐X‐C motif chemokine receptor 4 (CXCR4) in FS‐DFSP. Periostin‐positive CAFs coexpressed VEGFA and CCL2, potentially contributing to angiogenic features within the tumor microenvironment. These findings reveal that FS‐DFSP is characterized by coordinated angiogenic and stromal remodeling involving tumor cells and CAFs and suggest that the tumor microenvironment may represent a potential focus for future therapeutic investigation.

Keywords: cancer‐associated fibroblast, dermatofibrosarcoma protuberans, fibrosarcomatous dermatofibrosarcoma protuberans, tumor microenvironment, vascular endothelial growth factor A

FS‐DFSP exhibits coordinated angiogenic and stromal remodeling characterized by immature vascular features, CAF accumulation, and enhanced macrophage infiltration. VEGFA, CCL2, and CXCR4 may contribute to this tumor‐promoting microenvironment.

graphic file with name CAS-9999-0-g003.webp

Abbreviations

CAFs: cancer‐associated fibroblasts
CCL2: C‐C motif chemokine ligand 2
CCR2: C‐C motif chemokine receptor 2
COL1A1: collagen type 1 alpha 1
CXCL12: C‐X‐C motif chemokine ligand 12
CXCR4: C‐X‐C motif chemokine receptor 4
DEG: differentially expressed gene
DFSP: dermatofibrosarcoma protuberans
FAP: fibroblast activation protein
FFPE: formalin‐fixed paraffin‐embedded
FS: fibrosarcomatous
GSEA: gene set enrichment analysis
ICC: intraclass correlation coefficients
PDGFB: platelet‐derived growth factor B‐chain
PDGFR: platelet‐derived growth factor receptor
TME: tumor microenvironment
VEGFA: vascular endothelial growth factor A
αSMA: α‐smooth muscle actin

1. Introduction

Most patients with dermatofibrosarcoma protuberans (DFSP), an intermediate‐grade cutaneous sarcoma characterized by locally aggressive growth and frequent local recurrence [1], harbor the characteristic collagen type 1 alpha 1 (COL1A1) gene fused to platelet‐derived growth factor B‐chain (PDGFB). This event, resulting from the t(17;22)(q22;q13) chromosomal translocation, leads to the constitutive activation of platelet‐derived growth factor receptor (PDGFR) signaling [2, 3, 4, 5]. The World Health Organization classification recognizes several histologic DFSP variants, including Bednar tumor, giant cell fibroblastoma, and myxoid, myoid, and atrophic subtypes [1, 6]. DFSP may undergo fibrosarcomatous transformation, giving rise to fibrosarcomatous DFSP (FS‐DFSP) [7]. FS‐DFSP is characterized by increased cytologic atypia, elevated mitotic activity, and a higher risk of local recurrence and distant metastasis than pure‐DFSP [8, 9, 10], resulting in poorer clinical outcomes.

Tumor behavior is shaped not only by intrinsic genetic alterations but also by interactions between neoplastic cells and the tumor microenvironment (TME), including vascular, stromal, and immune components [11, 12]. Characterizing the TME is therefore important for understanding tumor progression, even in rare sarcomas such as DFSP.

Despite advances in the molecular pathology of DFSP, the genetic background of FS‐DFSP remains poorly understood due to its rarity. Recent single‐cell RNA sequencing studies have revealed that primary FS‐DFSP exhibits a pauci‐immune yet highly vascularized TME, with tumor cells contributing to angiogenesis through extracellular matrix remodeling and activation of transforming growth factor β and vascular endothelial growth factor (VEGF)‐related pathways [13]. These observations suggest that microenvironment‐associated factors, beyond constitutive COL1A1::PDGFB‐driven PDGFR signaling [2, 14, 15], contribute to intratumoral angiogenesis in FS‐DFSP. Moreover, the upregulation of angiogenesis‐ and stroma‐related gene signatures suggests that microenvironmental remodeling contributes to its aggressive behavior. Although tyrosine kinase inhibitors are available for unresectable DFSP, their efficacy remains limited in some patients, underscoring the need for further understanding of the TME and its therapeutic implications. However, how angiogenic and stromal features differ between pure‐DFSP and FS‐DFSP remains unclear. Therefore, this study aimed to compare the TME between pure‐DFSP and FS‐DFSP, focusing on angiogenic and stromal components, and identify pathologic features associated with tumor‐promoting mechanisms in FS‐DFSP.

2. Materials and Methods

2.1. Patients and Tumor Selection

The study included 122 patients diagnosed with DFSP whose tumor samples were archived between 1978 and 2023 at the Department of Anatomic Pathology in Kyushu University in Fukuoka, Japan. Diagnosis was based on histopathologic features, immunohistochemical CD34 positivity, or the detection of the COL1A1::PDGFB fusion gene, in accordance with the World Health Organization classification. Tumors without a fibrosarcomatous component were defined as pure‐DFSP, whereas tumors containing areas with a fascicular or herringbone growth pattern, increased cytologic atypia, and elevated mitotic activity were classified as FS‐DFSP. Formalin‐fixed paraffin‐embedded (FFPE) samples from 83 patients with pure‐DFSP and 39 with FS‐DFSP were analyzed (Table 1). None had received preoperative treatment, including chemotherapy or molecularly targeted therapy.

TABLE 1.

Clinicopathological characteristics of the 122 patients with DFSP.

Variables	Total (n = 122)	FS‐DFSP (n = 39)	Pure‐DFSP (n = 83)	p
Age mean (median, range)	44 (43, 6–82)	46 (44, 6–82)	43 (40, 9–81)	0.29
Sex male/female, n	67/55	24/15	43/40	0.31
Tumor site
Trunk, n (%)	97 (80)	32 (82)	65 (78)	0.63
Extremities, n (%)	19 (16)	6 (15)	13 (16)	0.96
Head and neck, n (%)	6 (5)	1 (3)	5 (6)	0.41

Open in a new tab

This retrospective study was approved by the Ethics Committee of Kyushu University (no. 23005) and conducted in accordance with the Declaration of Helsinki. The requirement for written informed consent was waived due to the retrospective study design, and an opt‐out procedure was implemented through public disclosure of the study information.

2.2. Evaluation of Intratumoral Microvessels

For comparative analyses, the evaluated tumor areas were categorized as DFSP (pure‐DFSP), DFSP (FS‐DFSP), and FS (FS‐DFSP), representing DFSP in pure‐DFSP, the DFSP component of FS‐DFSP, and the FS component of FS‐DFSP, respectively.

Intratumoral microvessels were assessed with immunostaining for CD31 and α‐smooth muscle actin (αSMA) in corresponding areas of the central tumor, excluding the tumor periphery and invasive front. CD31‐ and αSMA‐positive vascular abundance was independently graded as high, moderate, or low. Higher CD31 staining with lower αSMA staining was considered indicative of immature vascular structures [16, 17]. To validate the grading system, five cases per grade were randomly selected for each marker. CD31‐positive vascular area (%) and αSMA‐positive vessel density (vessels/mm²) were quantified using QuPath (version 0.7.0) [18], excluding nonvascular αSMA‐positive stromal regions, and compared with the corresponding semiquantitative grades.

2.3. RNA Extraction

Total RNA was extracted from the FFPE specimens using the RNAstorm FFPE RNA extraction kit (Cell Data Sciences, Fremont, CA, USA), according to the manufacturer's protocol. In pure‐DFSP cases, RNA was extracted from the DFSP region. In FS‐DFSP cases, FS (FS‐DFSP) and DFSP (FS‐DFSP) components were identified based on histologic features, including fascicular or herringbone architecture, increased cytologic atypia, and elevated mitotic activity. Reduced or absent CD34 immunoreactivity was used as an adjunct to distinguish the FS component from the adjacent DFSP component. RNA was extracted separately from each component.

2.4. Gene Expression Profiling and Bioinformatics Analysis

mRNA expression was assessed using the nCounter gene expression assay (NanoString Technologies) in 12 RNA samples: paired FS (FS‐DFSP) and DFSP (FS‐DFSP) components from 4 FS‐DFSP cases and DFSP (pure‐DFSP) samples from 4 pure‐DFSP cases. RNA samples were enriched for target regions using the nCounter Human Pan Cancer IO 360 panel and measured using the nCounter analysis system (NanoString Technologies). The raw data were normalized using nSolver analysis software (version 4.0, NanoString Technologies). The R package Rtsne was used to perform t‐distributed stochastic neighbor embedding. Gene set enrichment analysis (GSEA) was performed using the hallmark gene sets from the Molecular Signatures Database for three comparisons: (i) FS (FS‐DFSP) and DFSP (FS‐DFSP) combined versus DFSP (pure‐DFSP), (ii) FS (FS‐DFSP) versus DFSP (FS‐DFSP), and (iii) DFSP (FS‐DFSP) versus DFSP (pure‐DFSP) [19]. Differentially expressed genes (DEGs) with a log₂ fold change > 1.5 and a p‐value < 0.05 were identified using the R packages limma and edgeR [20, 21]. Paired FS (FS‐DFSP) versus DFSP (FS‐DFSP) comparisons were performed using limma with a paired design. Conversely, edgeR was used to compare FS (FS‐DFSP) and DFSP (FS‐DFSP) with DFSP (pure‐DFSP), which involved unpaired samples from different tumors.

2.5. Immunohistochemistry

Immunohistochemistry for CD34, CD31, αSMA, fibroblast activation protein (FAP), periostin, VEGF alpha (VEGFA), C‐C motif chemokine ligand 2 (CCL2), and CD163 was performed on FFPE sections using the primary antibodies listed in Table S1, as previously described [22]. Staining was independently reviewed by three pathologists (S.S., T.I., and Y.O.). CD34 immunostaining was considered positive when membranous or cytoplasmic staining was observed in tumor or endothelial cells. VEGFA and CCL2 immunostaining was considered positive if cytoplasmic staining was observed in tumor or stromal cells. CD31 immunostaining was considered positive if endothelial cells showed membranous staining. αSMA immunostaining was considered positive in the presence of cytoplasmic staining in perivascular cells or stromal myofibroblasts. Periostin and FAP immunostaining was considered positive in the presence of extracellular stromal deposition or cytoplasmic staining in stromal cells. CD163 immunostaining was considered positive in the presence of membranous or cytoplasmic staining in macrophages.

Because VEGFA and CCL2 are secreted proteins, immunohistochemistry was used primarily to evaluate their tissue localization and relative semiquantitative expression within each histological component [23, 24]. VEGFA and CCL2 expression levels were semiquantitatively evaluated using an H‐score method, calculated as the product of staining intensity (0–3) and the percentage of positive cells (0%–100%), yielding a score ranging from 0 to 300 [25, 26]. FAP and periostin immunoreactivity was semiquantitatively evaluated by multiplying staining intensity (0–3) by the estimated extent of positive stromal immunoreactivity (0%–100%), yielding a score ranging from 0 to 300 [27, 28]. To validate the semiquantitative FAP and periostin scores, five cases were randomly selected for each marker. In the tumor center and invasive front, intensity‐weighted digital scores were calculated using QuPath based on manually defined staining intensity thresholds and the corresponding positive stromal areas. CD163‐positive macrophages were manually counted in five high‐power fields obtained using a × 40 objective lens (field area: 0.237 mm²) in areas with the highest density, and the CD163‐positive macrophage density (cells/mm²) was calculated based on the total area of the five fields [29].

2.6. Immunofluorescence Staining

Immunofluorescence staining for CD34, periostin, VEGFA, and CCL2 was performed on FFPE tissue sections using standard protocols. Following heat‐induced antigen retrieval, sections were incubated with the respective primary antibodies (Table S1) and appropriate fluorophore‐conjugated secondary antibodies. Nuclei were counterstained with DAPI Counterstain (K004, GSP Laboratory, Japan), and images were acquired using a Leica DMi8 THUNDER fluorescence microscope system (Leica, Wetzlar, Germany). For CD34/periostin double immunofluorescence, CD34‐positive spindle cells were regarded as DFSP tumor cells, whereas spatially distinct periostin‐positive stromal cells were considered CAF‐associated stromal cells.

2.7. Statistical Analysis

All statistical analyses were performed using JMP Pro version 17.0.0 (SAS Institute, Cary, NC, USA) or R version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria). Continuous variables were compared using the Mann–Whitney U test, and categorical variables using the χ ² test or Fisher's exact test, as appropriate. Comparisons between FS (FS‐DFSP) and DFSP (FS‐DFSP) from the same FS‐DFSP cases were performed using the Wilcoxon signed‐rank test.

Immunohistochemical scores were analyzed as continuous variables. Expression levels of VEGFA, CCL2, and CXCR4 were analyzed using the limma package in R software. Patient‐matched FS (FS‐DFSP) and DFSP (FS‐DFSP) components from the same FS‐DFSP cases were modeled as paired samples. Interobserver reliability for the evaluation of intratumoral microvessels based on CD31 and αSMA immunostaining was assessed using intraclass correlation coefficients (ICCs) in JMP Pro. The ICC value corresponding to “within‐class correlation including bias and interaction” was used as an indicator of reliability. Correlations between semiquantitative assessments and digital measurements were evaluated using Spearman's rank correlation coefficient. All statistical tests were two‐sided, and p‐values < 0.05 were considered statistically significant.

3. Results

3.1. Clinicopathological Characteristics

The clinicopathologic features of the study cohort are summarized in Table 1. The cohort comprised 122 patients with DFSP, including 67 males and 55 females, with a median age of 44 years (range, 6–82 years). Of these, 83 had pure‐DFSP and 39 had FS‐DFSP. The trunk was the most frequent tumor site (n = 97, 79.5%), followed by the extremities (n = 19, 15.6%) and head and neck (n = 6, 4.9%). Demographic and anatomic characteristics were comparable between the two groups, with no significant differences in age, sex, or tumor site (Table 1).

3.2. Evaluation of Intratumoral Vasculature and Angiogenesis in FS‐DFSP

CD31‐ and αSMA‐positive vascular abundance was assessed in DFSP (pure‐DFSP), DFSP (FS‐DFSP), and FS (FS‐DFSP) (Figure 1; Figure S1). CD31‐positive vascular abundance differed significantly among the three groups (p < 0.0001). High, moderate, and low CD31 staining levels were observed in 16.87%, 38.55%, and 44.58% of DFSP (pure‐DFSP); 38.46%, 20.51%, and 41.03% of DFSP (FS‐DFSP); and 64.10%, 17.95%, and 17.95% of FS (FS‐DFSP), respectively. In contrast, αSMA‐positive vascular abundance did not differ significantly (p = 0.2090). High, moderate, and low αSMA staining levels were observed in 30.12%, 28.92%, and 40.96% of DFSP (pure‐DFSP); 33.33%, 35.90%, and 30.77% of DFSP (FS‐DFSP); and 46.15%, 33.33%, and 20.51% of FS (FS‐DFSP), respectively. These findings indicate increased CD31‐positive vascular abundance in FS (FS‐DFSP), whereas αSMA‐positive vascular structures did not differ significantly among the three groups, reflecting an increased abundance of immature vascular structures in FS (FS‐DFSP). Interobserver agreement was excellent, with ICC (1, 2) of 0.93 and 0.91 for CD31 and αSMA, respectively. Digital measurements correlated strongly with the semiquantitative grades for CD31 (Spearman's ρ = 0.91, p < 0.001) and αSMA (ρ = 0.87, p < 0.001) (Figure S2).

Histopathological features and intratumoral vasculature in FS‐DFSP and pure‐DFSP. Representative images of FS (FS‐DFSP), DFSP (FS‐DFSP), and DFSP (pure‐DFSP) are shown. The left column shows FS (FS‐DFSP) (a, d, g, j), the middle column DFSP (FS‐DFSP) (b, e, h, k), and the right column DFSP (pure‐DFSP) (c, f, i, l). H&E staining demonstrates the characteristic morphology of each component (a–c). CD34 staining (d–f) shows diffuse positivity in DFSP (FS‐DFSP) and DFSP (pure‐DFSP), with loss or reduction of expression in FS (FS‐DFSP). Intratumoral vasculature was assessed by CD31 (g–i) and αSMA (j–l) immunostaining. CD31 and αSMA staining was performed on serial sections of corresponding tumor areas, allowing comparison of vascular structures between the two markers. Increased CD31‐positive vessels without a corresponding increase in αSMA‐positive vessels in FS (FS‐DFSP) reflect an increase in immature vasculature. Scale bars: 50 μm.

3.3. Evaluation of Cancer‐Associated Fibroblasts in FS‐DFSP

Immunohistochemical staining for the cancer‐associated fibroblast (CAF) markers FAP and periostin revealed significantly higher semiquantitative stromal immunoreactivity scores in the FS‐DFSP group than in the pure‐DFSP group for both the intratumoral region and the invasive front of the tumor (p < 0.001 for both; Mann–Whitney U test) (Figure 2A). The median intratumoral FAP scores were 0 (range, 0–80) and 8.6 (range, 0–150) in the pure‐DFSP and FS‐DFSP groups, respectively, whereas those at the invasive front were 3.0 (range, 0–155) and 19 (range, 1–148), respectively. Similarly, the median intratumoral periostin scores were 1.6 (range, 0–108) and 50 (range, 0–146), respectively, whereas those at the invasive front of the tumor were 20 (range, 0–138) and 66 (range, 0–154), respectively. Both markers showed spatially heterogeneous stromal immunoreactivity, with accentuation at the invasive front. Scores were significantly higher at the invasive front than in the intratumoral region, particularly in FS‐DFSP (p < 0.001). Digital scores showed strong positive associations with the semiquantitative scores for periostin in both regions and for FAP in the tumor center. The association for FAP at the invasive front was also positive but did not reach statistical significance (ρ = 0.82, p = 0.089) (Figure S3). These findings suggested preferential accumulation of CAF‐associated stromal components at the invasive front of FS‐DFSP. αSMA showed partially overlapping staining patterns with FAP and periostin but was also focally detected in cells morphologically consistent with tumor cells.

FAP‐ and periostin‐expressing CAFs in FS‐DFSP. (A) Representative H&E and immunohistochemical images of FAP and periostin in FS‐DFSP. The left column shows the FS component in the tumor center, and the right column shows the invasive front of the same tumor. FAP and periostin staining demonstrate the distribution of FAP‐ and periostin‐positive stromal cells within the tumor and at the invasive front. (B) Representative H&E and double immunofluorescence for CD34 (green), periostin (red), and DAPI (blue) in the CD34‐positive DFSP component of FS‐DFSP. The dotted line in the H&E image indicates the boundary between the tumor and stroma. CD34‐positive tumor cells () were spatially distinct from periostin‐positive stromal cells (), supporting the presence of a CAF‐associated stromal population distinct from tumor cells. Scale bars: 50 μm (A, left); 200 μm (A, right; B, left and middle); 5 μm (B, right).

Inline graphic — FAP‐ and periostin‐expressing CAFs in FS‐DFSP. (A) Representative H&E and immunohistochemical images of FAP and periostin in FS‐DFSP. The left column shows the FS component in the tumor center, and the right column shows the invasive front of the same tumor. FAP and periostin staining demonstrate the distribution of FAP‐ and periostin‐positive stromal cells within the tumor and at the invasive front. (B) Representative H&E and double immunofluorescence for CD34 (green), periostin (red), and DAPI (blue) in the CD34‐positive DFSP component of FS‐DFSP. The dotted line in the H&E image indicates the boundary between the tumor and stroma. CD34‐positive tumor cells () were spatially distinct from periostin‐positive stromal cells (), supporting the presence of a CAF‐associated stromal population distinct from tumor cells. Scale bars: 50 μm (A, left); 200 μm (A, right; B, left and middle); 5 μm (B, right).

Double immunofluorescence staining for CD34 and periostin demonstrated spatial separation between CD34‐positive DFSP tumor cells and periostin‐positive stromal cells (Figure 2B), supporting the presence of a periostin‐positive stromal population distinct from the tumor cells.

3.4. mRNA Expression Analysis

mRNA expression was analyzed across the three histologically defined components (Figure 1). t‐Distributed stochastic neighbor embedding (t‐SNE) analysis showed that DFSP (FS‐DFSP), despite minimal histologic atypia, clustered more closely with FS (FS‐DFSP) than with histologically similar DFSP (pure‐DFSP). DFSP (pure‐DFSP) samples formed a cluster distinct from the FS‐DFSP samples. These findings suggest that TME‐associated transcriptomic alterations shared with the fibrosarcomatous component may precede overt fibrosarcomatous morphology in FS‐DFSP (Figure 3).

Transcriptomic profiles of FS and DFSP components. t‐SNE visualization of gene expression profiles in FS (FS‐DFSP), DFSP (FS‐DFSP), and DFSP (pure‐DFSP). FS (FS‐DFSP) and DFSP (FS‐DFSP) showed relatively similar transcriptomic profiles, whereas DFSP (pure‐DFSP) formed a more distinct cluster.

Table S2 shows the GSEA results for the three groups using hallmark gene sets from the Molecular Signatures Database. In the comparison of the combined FS (FS‐DFSP) and DFSP (FS‐DFSP) group with DFSP (pure‐DFSP), HALLMARK_INTERFERON_ALPHA_RESPONSE and HALLMARK_INTERFERON_GAMMA_RESPONSE were significantly enriched in the combined FS (FS‐DFSP) and DFSP (FS‐DFSP) group, indicating the activation of interferon‐related immune programs in FS‐DFSP‐associated components.

The comparison between FS (FS‐DFSP) and DFSP (FS‐DFSP) revealed that gene sets related to cell‐cycle regulation and proliferative activity, including HALLMARK_E2F_TARGETS and HALLMARK_G2M_CHECKPOINT, were consistently enriched in FS (FS‐DFSP). Although these differences did not reach statistical significance after correction for multiple testing, the concordant upregulation of multiple cell cycle‐associated programs suggested a trend toward enhanced proliferative signaling in FS (FS‐DFSP).

The comparison of DFSP (FS‐DFSP) with DFSP (pure‐DFSP) revealed that HALLMARK_INTERFERON_ALPHA_RESPONSE and HALLMARK_INTERFERON_GAMMA_RESPONSE were significantly enriched in DFSP (FS‐DFSP), along with nominal enrichment of HALLMARK_IL6_JAK_STAT3_SIGNALING. Collectively, interferon‐response signatures were the most robustly enriched programs in FS‐DFSP‐associated samples compared with DFSP (pure‐DFSP), whereas FS (FS‐DFSP) showed a trend toward increased cell‐cycle activity compared with DFSP (FS‐DFSP). The leading‐edge genes contributing to these enriched hallmark gene sets included CCL2 and CXCR4, both of which have been implicated in angiogenesis and tumor microenvironment interactions (Table S2).

3.5. Upregulation of Angiogenesis‐Related Genes in FS‐DFSP

Integration of GSEA and DEG analysis (Tables S2 and S3) highlighted CCL2 and CXCR4 as candidate molecules associated with intratumoral angiogenesis and fibrosarcomatous progression. CCL2, a leading‐edge gene within the enriched interferon‐related hallmark gene sets, was upregulated in FS‐DFSP–associated components (Table S2). In contrast, DEG analysis revealed that CXCR4 was upregulated in FS (FS‐DFSP) compared with DFSP (pure‐DFSP) (Table S3).

Although VEGFA was neither a significant DEG nor enriched in GSEA, VEGFA was selected for further analysis based on the immature vasculature and vascular remodeling observed in FS‐DFSP. We therefore examined VEGFA, CCL2, and CXCR4 mRNA expression across the three tumor components. VEGFA expression was significantly higher in FS (FS‐DFSP) than in DFSP (pure‐DFSP) (p = 0.049). In contrast, both FS (FS‐DFSP) and DFSP (FS‐DFSP) showed significantly higher CCL2 expression than DFSP (pure‐DFSP) (p = 0.033 and p = 0.040, respectively), whereas FS (FS‐DFSP) and DFSP (FS‐DFSP) did not differ significantly (p = 0.875). Although CXCR4 expression did not differ significantly among the groups, it tended to be higher in FS (FS‐DFSP) than in DFSP (FS‐DFSP) and DFSP (pure‐DFSP) (Figure 4A).

Expression of angiogenesis‐related genes and proteins. (A) Expression of angiogenesis‐related genes (VEGFA, CCL2, and CXCR4). All three genes tended to be upregulated in FS (FS‐DFSP). (B) Representative immunohistochemical images of VEGFA and CCL2 showing their immunoreactivity within the tumor. Scale bars: 50 μm. (C) Semiquantitative H‐score analysis of VEGFA and CCL2 immunoreactivity in FS (FS‐DFSP), DFSP (FS‐DFSP), and DFSP (pure‐DFSP). Comparisons were performed using the Wilcoxon signed‐rank test or Mann–Whitney U test, as appropriate. p‐values are indicated in the figure.

To further examine the transcriptomic findings, immunoreactivity for VEGFA and CCL2 was assessed in the three tumor components (Figure 4B,C). Both markers showed stronger tumor cell immunoreactivity in FS (FS‐DFSP) than in DFSP (FS‐DFSP) or DFSP (pure‐DFSP) (Figure 4B). Semiquantitative H‐score analysis showed significantly higher VEGFA immunoreactivity in FS (FS‐DFSP) than in both DFSP (FS‐DFSP) and DFSP (pure‐DFSP) (both p < 0.001), whereas no significant difference was observed between the latter two groups (p = 0.379). Similarly, CCL2 immunoreactivity was significantly higher in FS (FS‐DFSP) than in both DFSP (FS‐DFSP) (p < 0.001) and DFSP (pure‐DFSP) (p = 0.005). While the difference between DFSP (FS‐DFSP) and DFSP (pure‐DFSP) was not statistically significant (p = 0.086), DFSP (FS‐DFSP) showed a trend toward higher CCL2 immunoreactivity than DFSP (pure‐DFSP) (Figure 4C). These findings were generally consistent with the transcriptomic results, supporting increased VEGFA and CCL2 expression in FS‐DFSP.

3.6. Evaluation of Macrophage Infiltration in FS‐DFSP

Macrophage infiltration was evaluated using CD163 immunohistochemistry as a marker associated with M2‐like macrophages (Figure 5). Quantitative analysis of CD163‐positive macrophages (cells/mm²) revealed significantly higher macrophage density in FS (FS‐DFSP) than in both DFSP (FS‐DFSP) (p < 0.001) and DFSP (pure‐DFSP) (p = 0.003) (Figure 5A). Although not statistically significant, macrophage density tended to be higher in DFSP (FS‐DFSP) than in DFSP (pure‐DFSP) (p = 0.089). Consistent with these findings, immunohistochemistry demonstrated abundant CD163‐positive macrophage infiltration in FS (FS‐DFSP), whereas fewer CD163‐positive macrophages were observed in DFSP (FS‐DFSP) and DFSP (pure‐DFSP) (Figure 5B).

CD163‐positive macrophage infiltration in FS and DFSP components. (A) Quantification of CD163‐positive macrophages in FS (FS‐DFSP), DFSP (FS‐DFSP), and DFSP (pure‐DFSP). CD163‐positive macrophage density was significantly higher in FS (FS‐DFSP) than in the other two groups. Comparisons were performed using the Wilcoxon signed‐rank test or Mann–Whitney U test, as appropriate. p‐values are indicated in the figure. (B) Representative immunohistochemical images of CD163 in FS (FS‐DFSP), DFSP (FS‐DFSP), and DFSP (pure‐DFSP). Scale bars: 50 μm.

3.7. Expression of VEGFA and CCL2 in Periostin‐Positive CAFs

Immunofluorescence staining revealed that periostin‐positive CAFs exhibited cytoplasmic staining for VEGFA and CCL2, indicating that CAFs might contribute to the TME through the expression of pro‐angiogenic and inflammatory factors (Figure S4).

4. Discussion

In the present study, we performed a comprehensive comparison of the TME between FS‐DFSP and pure‐DFSP using histopathologic and transcriptomic approaches. FS‐DFSP was characterized by an angiogenic and stromal microenvironment marked by increased angiogenesis with immature vascular features, accumulation of CAFs, and enhanced macrophage infiltration. Transcriptomic profiling revealed shared molecular features between DFSP (FS‐DFSP) and FS (FS‐DFSP), suggesting that progression‐associated alterations may precede overt fibrosarcomatous morphology.

Gene expression analyses demonstrated the upregulation of angiogenesis‐related molecules, including VEGFA, CCL2, and CXCR4, in FS (FS‐DFSP). Although PDGFR signaling driven by the COL1A1::PDGFB fusion protein has long been considered a central mechanism of tumor growth and angiogenesis in DFSP [2, 14, 15], our findings suggest that other TME‐associated factors and molecular pathways may contribute to enhanced angiogenesis in FS‐DFSP. Increased intratumoral angiogenesis is associated with aggressive tumor behavior [30, 31], and consistent with this, increased angiogenesis in FS‐DFSP may be a characteristic feature of its aggressive phenotype.

VEGFA promotes endothelial cell proliferation and migration, leading to immature vessels and tumor progression [32, 33, 34]. Although neither DEG analysis nor GSEA identified significant VEGFA‐related enrichment, transcriptomic and immunohistochemical analyses supported increased VEGFA expression in FS (FS‐DFSP), consistent with its immature vascular features. CCL2 promotes tumor progression by recruiting CCR2‐positive M2‐like tumor‐associated macrophages and directly promoting angiogenesis, which is associated with aggressive tumor behavior and poor clinical outcomes, while CCL2 neutralization can suppress angiogenic activity and metastatic potential [35, 36, 37]. CXCR4 regulates hematopoietic and bone marrow‐derived cell trafficking through CXCL12 signaling [38]. The CXCL12–CXCR4 axis promotes endothelial progenitor cell mobilization and cooperates with VEGFA to enhance pathologic angiogenesis [39], potentially contributing to immature vascular features in FS‐DFSP.

M2‐like macrophages promote tumor progression by enhancing angiogenesis, invasion, and immunosuppression [35]. We observed increased infiltration of CD163‐positive macrophages in FS (FS‐DFSP). CCL2 promotes M2‐like macrophage recruitment, while VEGFA has also been implicated in M2‐like polarization; VEGFA knockdown reportedly inhibits this process [37]. CXCL12–CXCR4 signaling also promotes M2‐like macrophage infiltration [40]. Thus, CD163‐positive macrophages may contribute to the angiogenic microenvironment of FS‐DFSP.

We identified FAP‐ and periostin‐positive CAFs distinct from tumor cells. These CAF‐associated markers were increased in FS‐DFSP compared with pure‐DFSP, suggesting that CAF activation is a characteristic stromal feature of FS‐DFSP. CAFs are heterogeneous, and αSMA is a marker of myofibroblastic CAFs (mCAFs) [41]; however, its heterogeneous stromal expression and focal positivity in morphologically suspected tumor cells in DFSP precluded its use for CAF subtyping.

Immunoreactivity for VEGFA and CCL2 was observed in association with periostin‐positive stromal cells (Figure S4), suggesting that stromal cells, in addition to tumor cells, may contribute to the proangiogenic microenvironment. Activated CAFs may also amplify CCL2–CCR2 and CXCL12–CXCR4 signaling [42, 43], promoting angiogenesis and stromal remodeling.

In recent years, periostin‐targeted therapies have been explored, with anti‐periostin antibodies suppressing metastasis in breast cancer models [44, 45]. The preferential accumulation of CAFs at the invasive front may suggest their involvement in local tumor progression and invasion. Given the infiltrative growth of DFSP and the risk of positive surgical margins, stromal periostin may warrant investigation as a potential neoadjuvant therapeutic target to improve local control.

Component‐based differential expression analysis revealed stage‐dependent angiogenic programs across the three tumor regions. CCL2 was upregulated in DFSP (FS‐DFSP) compared with DFSP (pure‐DFSP), suggesting early activation of the CCL2 axis before overt fibrosarcomatous morphology. In contrast, FS (FS‐DFSP) showed increases in VEGFA and CXCR4, suggesting further progressive angiogenic and microenvironmental remodeling during fibrosarcomatous transformation. These findings support the biological relevance of analyzing distinct components and suggest coordinated microenvironmental remodeling underlying the angiogenic phenotype of FS‐DFSP. In addition, CCL2 upregulation in the DFSP (FS‐DFSP) suggests its potential utility as a diagnostic marker for distinguishing FS‐DFSP from pure‐DFSP.

Overall, our findings suggest that targeting angiogenesis‐related and CAF‐associated pathways alongside PDGFR inhibition may offer therapeutic opportunities in FS‐DFSP, although further clinical investigation is needed.

This study has several limitations. The transcriptomic analysis involved a small subset because obtaining paired components with adequate RNA quality from archival FFPE tissues was challenging. In addition, the use of a targeted expression panel may have limited the identification of additional pathways relevant to fibrosarcomatous transformation. Furthermore, immunohistochemical and immunofluorescence analyses used selected stromal and immune markers, providing a focused rather than comprehensive characterization of the TME. Digital vascular assessment was limited by interconnected CD31‐positive vessels, which hindered vessel counting, and variations in vascular wall thickness, which affected αSMA‐positive area measurements and required different quantification methods. Nonvascular αSMA‐positive stromal regions required manual exclusion, while CAF assessment required manual ROI selection based on histopathology to avoid including non‐tumor regions, highlighting the limitations of digital image analysis in soft tissue tumors. Although double immunofluorescence supported the distinction between CD34‐positive tumor cells and periostin‐positive stromal cells, reduced CD34 expression in fibrosarcomatous areas may limit this approach. Immunohistochemical staining of secreted VEGFA and CCL2 may not directly reflect absolute protein levels, although transcriptomic findings supported their expression patterns. Owing to the rarity of DFSP and the limited number of metastatic or recurrent events in our cohort, robust survival analyses could not be performed. Therefore, the prognostic significance of the vascular and stromal alterations remains unclear. Future functional and clinical studies in larger cohorts are warranted to establish the mechanistic and prognostic significance of VEGFA‐, CCL2‐, and CXCR4‐associated microenvironmental remodeling in FS‐DFSP.

FS‐DFSP is characterized by an angiogenic and stromal TME with immature vascular features, CAF accumulation, and enhanced macrophage infiltration. Our findings suggest that VEGFA, CCL2, and CXCR4 are involved in the angiogenic phenotype of FS‐DFSP, with potential contributions from M2‐like tumor‐associated macrophages and periostin‐positive CAFs. These observations provide new insights into vascular and stromal remodeling in FS‐DFSP and may inform future therapeutic strategies targeting angiogenic pathways.

Author Contributions

Sakura Shiraishi: conceptualization, investigation, writing – original draft, methodology, visualization, data curation, formal analysis. Takeshi Iwasaki: conceptualization, methodology, supervision, writing – review and editing. Ken Takigawa: methodology, investigation. Kengo Kawaguchi: resources. Hiroki Sonoda: resources. Kohta Miyawaki: methodology, investigation. Takumi Tomonaga: resources. Taro Mori: resources. Makoto Endo: resources. Koji Kato: resources, supervision. Yasuharu Nakashima: supervision, resources. Yoshinao Oda: conceptualization, supervision, funding acquisition, project administration.

Funding

This work was supported by the Japan Society for the Promotion of Science (Nos. 23 K27385 Y.O.).

Ethics Statement

Approved by the Ethics Committee of Kyushu University (no. 23005).

Consent

The authors have nothing to report.

Conflicts of Interest

Yoshinao Oda is an editorial board member of Cancer Science. The other authors declare no conflicts of interest.

Supporting information

Figure S1: Semiquantitative evaluation of vascular markers.

Figure S2: Digital validation of semiquantitative vascular grading.

Figure S3: Digital validation of semiquantitative periostin and FAP scores.

Figure S4: Immunofluorescence staining of CAF‐derived VEGFA and CCL2.

Table S1: List of primary antibodies.

Table S2: Gene set enrichment analysis (GSEA) of Hallmark pathways across distinct tumor components in FS (FS‐DFSP), DFSP (FS‐DFSP), and DFSP (pure‐DFSP).

Table S3: Differentially expressed genes among FS (FS‐DFSP), DFSP (FS‐DFSP), and DFSP (pure‐DFSP).

CAS-9999-0-s001.pdf^{(696.8KB, pdf)}

Acknowledgments

The authors would like to thank Enago for the English language review. We thank Motoko Tomita, Naoko Ieiri‐Otani, Haruka Toki, and Jumi Yahiro at the Department of Anatomic Pathology for providing technical support.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

References

1. WHO Classification of Tumours Editorial Board , Soft Tissue and Bone Tumours, 5th ed. (International Agency for Research on Cancer, 2020). [Google Scholar]
2. Simon M. P., Pedeutour F., Sirvent N., et al., “Deregulation of the Platelet‐Derived Growth Factor B‐Chain Gene via Fusion With Collagen Gene COL1A1 in Dermatofibrosarcoma Protuberans and Giant‐Cell Fibroblastoma,” Nature Genetics 15, no. 1 (1997): 95–98. [DOI] [PubMed] [Google Scholar]
3. Sirvent N., Maire G., and Pedeutour F., “Genetics of Dermatofibrosarcoma Protuberans Family of Tumors: From Ring Chromosomes to Tyrosine Kinase Inhibitor Treatment,” Genes, Chromosomes & Cancer 37, no. 1 (2003): 1–19. [DOI] [PubMed] [Google Scholar]
4. Takahira T., Oda Y., Tamiya S., et al., “Detection of COL1A1‐PDGFB Fusion Transcripts and PDGFB/PDGFRB mRNA Expression in Dermatofibrosarcoma Protuberans,” Modern Pathology 20, no. 6 (2007): 668–675. [DOI] [PubMed] [Google Scholar]
5. Patel K. U., Szabo S. S., Hernandez V. S., et al., “Dermatofibrosarcoma Protuberans COL1A1‐PDGFB Fusion Is Identified in Virtually All Dermatofibrosarcoma Protuberans Cases When Investigated by Newly Developed Multiplex Reverse Transcription Polymerase Chain Reaction and Fluorescence in Situ Hybridization Assays,” Human Pathology 39, no. 2 (2008): 184–193. [DOI] [PubMed] [Google Scholar]
6. Iwasaki T., Yamamoto H., and Oda Y., “Current Update on the Molecular Biology of Cutaneous Sarcoma: Dermatofibrosarcoma Protuberans,” Current Treatment Options in Oncology 20, no. 4 (2019): 29. [DOI] [PubMed] [Google Scholar]
7. Mentzel T., Beham A., Katenkamp D., Dei Tos A. P., and Fletcher C. D., “Fibrosarcomatous (“High‐Grade”) Dermatofibrosarcoma Protuberans: Clinicopathologic and Immunohistochemical Study of a Series of 41 Cases With Emphasis on Prognostic Significance,” American Journal of Surgical Pathology 22, no. 5 (1998): 576–587. [DOI] [PubMed] [Google Scholar]
8. Goldblum J. R., Reith J. D., and Weiss S. W., “Sarcomas Arising in Dermatofibrosarcoma Protuberans: A Reappraisal of Biologic Behavior in Eighteen Cases Treated by Wide Local Excision With Extended Clinical Follow Up,” American Journal of Surgical Pathology 24, no. 8 (2000): 1125–1130. [DOI] [PubMed] [Google Scholar]
9. Bowne W. B., Antonescu C. R., Leung D. H., et al., “Dermatofibrosarcoma Protuberans: A Clinicopathologic Analysis of Patients Treated and Followed at a Single Institution,” Cancer 88, no. 12 (2000): 2711–2720. [PubMed] [Google Scholar]
10. Abbott J. J., Oliveira A. M., and Nascimento A. G., “The Prognostic Significance of Fibrosarcomatous Transformation in Dermatofibrosarcoma Protuberans,” American Journal of Surgical Pathology 30, no. 4 (2006): 436–443. [DOI] [PubMed] [Google Scholar]
11. Hanahan D. and Weinberg R. A., “Hallmarks of Cancer: The Next Generation,” Cell 144, no. 5 (2011): 646–674. [DOI] [PubMed] [Google Scholar]
12. Quail D. F. and Joyce J. A., “Microenvironmental Regulation of Tumor Progression and Metastasis,” Nature Medicine 19, no. 11 (2013): 1423–1437. [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Peng R., Li Y., Gao Y., Chen D., Li Z., and Zhao Y., “Cellular and Molecular Landscape of Primary Dermatofibrosarcoma Protuberans: Insights From Single‐Cell RNA Sequencing Analysis,” Experimental Dermatology 33, no. 8 (2024): e15121. [DOI] [PubMed] [Google Scholar]
14. Linn S. C., West R. B., Pollack J. R., et al., “Gene Expression Patterns and Gene Copy Number Changes in Dermatofibrosarcoma Protuberans,” American Journal of Pathology 163, no. 6 (2003): 2383–2395. [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Zou X., Tang X. Y., Qu Z. Y., et al., “Targeting the PDGF/PDGFR Signaling Pathway for Cancer Therapy: A Review,” International Journal of Biological Macromolecules 202 (2022): 539–557. [DOI] [PubMed] [Google Scholar]
16. Saijo Y., Furumoto H., Yoshida K., Nishimura M., and Irahara M., “Clinical Significance of Vascular Endothelial Growth Factor Expression and Microvessel Density in Invasive Cervical Cancer,” Journal of Medical Investigation 62, no. 3–4 (2015): 154–160. [DOI] [PubMed] [Google Scholar]
17. Usuda K., Iwai S., Funasaki A., et al., “Expression and Prognostic Impact of VEGF, CD31 and alphaSMA in Resected Primary Lung Cancers,” Anticancer Research 38, no. 7 (2018): 4057–4063. [DOI] [PubMed] [Google Scholar]
18. Bankhead P., Loughrey M. B., Fernandez J. A., et al., “QuPath: Open Source Software for Digital Pathology Image Analysis,” Scientific Reports 7, no. 1 (2017): 16878. [DOI] [PMC free article] [PubMed] [Google Scholar]
19. Subramanian A., Tamayo P., Mootha V. K., et al., “Gene Set Enrichment Analysis: A Knowledge‐Based Approach for Interpreting Genome‐Wide Expression Profiles,” Proceedings of the National Academy of Sciences of the United States of America 102, no. 43 (2005): 15545–15550. [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Ritchie M. E., Phipson B., Wu D., et al., “Limma Powers Differential Expression Analyses for RNA‐Sequencing and Microarray Studies,” Nucleic Acids Research 43, no. 7 (2015): e47. [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Robinson M. D., McCarthy D. J., and Smyth G. K., “edgeR: A Bioconductor Package for Differential Expression Analysis of Digital Gene Expression Data,” Bioinformatics 26, no. 1 (2010): 139–140. [DOI] [PMC free article] [PubMed] [Google Scholar]
22. Sonoda H., Iwasaki T., Ishihara S., Mori T., Nakashima Y., and Oda Y., “Impact of Tertiary Lymphoid Structure on Prognosis and Tumor Microenvironment in Undifferentiated Pleomorphic Sarcoma,” Cancer Science 116, no. 5 (2025): 1464–1473. [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Schirizzi A., Arshadi A., Tolomeo D., et al., “VEGFA Status as a Predictive Marker of Therapy Outcome in Metastatic Gastric Cancer Patients Following Ramucirumab‐Based Treatment,” Biomedicines 11, no. 10 (2023): 2721. [DOI] [PMC free article] [PubMed] [Google Scholar]
24. Eckstein M., Epple E., Jung R., et al., “CCL2 Expression in Tumor Cells and Tumor‐Infiltrating Immune Cells Shows Divergent Prognostic Potential for Bladder Cancer Patients Depending on Lymph Node Stage,” Cancers (Basel) 12, no. 5 (2020): 1253. [DOI] [PMC free article] [PubMed] [Google Scholar]
25. K. S. McCarty, Jr. , Miller L. S., Cox E. B., Konrath J., and K. S. McCarty, Sr. , “Estrogen Receptor Analyses. Correlation of Biochemical and Immunohistochemical Methods Using Monoclonal Antireceptor Antibodies,” Archives of Pathology & Laboratory Medicine 109, no. 8 (1985): 716–721. [PubMed] [Google Scholar]
26. Wang G., Zhang H., Shen X., Jin W., Wang X., and Zhou Z., “Characterization of Cancer‐Associated Fibroblasts (CAFs) and Development of a CAF‐Based Risk Model for Triple‐Negative Breast Cancer,” Cancer Cell International 23, no. 1 (2023): 294. [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Tchou J., Zhang P. J., Bi Y., et al., “Fibroblast Activation Protein Expression by Stromal Cells and Tumor‐Associated Macrophages in Human Breast Cancer,” Human Pathology 44, no. 11 (2013): 2549–2557. [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Jeon Y., Lee G., Jeong B. K., et al., “Proteomic Analysis Identifies Association of Periostin, a Matricellular Protein, With High Tumor Stroma and Immune Exclusion in Triple‐Negative Breast Cancer,” Cancer Research and Treatment 58, no. 3 (2026): 748–769. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Perea F., Bernal M., Sanchez‐Palencia A., et al., “The Absence of HLA Class I Expression in Non‐Small Cell Lung Cancer Correlates With the Tumor Tissue Structure and the Pattern of T Cell Infiltration,” International Journal of Cancer 140, no. 4 (2017): 888–899. [DOI] [PubMed] [Google Scholar]
30. Weidner N., “Tumor Angiogenesis: Review of Current Applications in Tumor Prognostication,” Seminars in Diagnostic Pathology 10, no. 4 (1993): 302–313. [PubMed] [Google Scholar]
31. Lugano R., Ramachandran M., and Dimberg A., “Tumor Angiogenesis: Causes, Consequences, Challenges and Opportunities,” Cellular and Molecular Life Sciences 77, no. 9 (2020): 1745–1770. [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Ferrara N., “VEGF as a Therapeutic Target in Cancer,” Oncology 69, no. Suppl 3 (2005): 11–16. [DOI] [PubMed] [Google Scholar]
33. Apte R. S., Chen D. S., and Ferrara N., “VEGF in Signaling and Disease: Beyond Discovery and Development,” Cell 176, no. 6 (2019): 1248–1264. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Zhang L., Zhang Y., Li X., Gao H., Chen X., and Li P., “CircRNA‐miRNA‐VEGFA: An Important Pathway to Regulate Cancer Pathogenesis,” Frontiers in Pharmacology 14 (2023): 1049742. [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Qian B. Z. and Pollard J. W., “Macrophage Diversity Enhances Tumor Progression and Metastasis,” Cell 141, no. 1 (2010): 39–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
36. Xu M., Wang Y., Xia R., Wei Y., and Wei X., “Role of the CCL2‐CCR2 Signalling Axis in Cancer: Mechanisms and Therapeutic Targeting,” Cell Proliferation 54, no. 10 (2021): e13115. [DOI] [PMC free article] [PubMed] [Google Scholar]
37. Ren Z., Xu Z., Chang X., Liu J., and Xiao W., “STC1 Competitively Binding betaPIX Enhances Melanoma Progression via YAP Nuclear Translocation and M2 Macrophage Recruitment Through the YAP/CCL2/VEGFA/AKT Feedback Loop,” Pharmacological Research 204 (2024): 107218. [DOI] [PubMed] [Google Scholar]
38. Teicher B. A. and Fricker S. P., “CXCL12 (SDF‐1)/CXCR4 Pathway in Cancer,” Clinical Cancer Research 16, no. 11 (2010): 2927–2931. [DOI] [PubMed] [Google Scholar]
39. Liang Z., Brooks J., Willard M., et al., “CXCR4/CXCL12 Axis Promotes VEGF‐Mediated Tumor Angiogenesis Through Akt Signaling Pathway,” Biochemical and Biophysical Research Communications 359, no. 3 (2007): 716–722. [DOI] [PMC free article] [PubMed] [Google Scholar]
40. Beider K., Bitner H., Leiba M., et al., “Multiple Myeloma Cells Recruit Tumor‐Supportive Macrophages Through the CXCR4/CXCL12 Axis and Promote Their Polarization Toward the M2 Phenotype,” Oncotarget 5, no. 22 (2014): 11283–11296. [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Sahai E., Astsaturov I., Cukierman E., et al., “A Framework for Advancing Our Understanding of Cancer‐Associated Fibroblasts,” Nature Reviews. Cancer 20, no. 3 (2020): 174–186. [DOI] [PMC free article] [PubMed] [Google Scholar]
42. Liang T., Tao T., Wu K., et al., “Cancer‐Associated Fibroblast‐Induced Remodeling of Tumor Microenvironment in Recurrent Bladder Cancer,” Advanced Science (Weinheim, Baden‐Wurttemberg, Germany) 10, no. 31 (2023): e2303230. [DOI] [PMC free article] [PubMed] [Google Scholar]
43. Kalluri R., “The Biology and Function of Fibroblasts in Cancer,” Nature Reviews. Cancer 16, no. 9 (2016): 582–598. [DOI] [PubMed] [Google Scholar]
44. Kyutoku M., Taniyama Y., Katsuragi N., et al., “Role of Periostin in Cancer Progression and Metastasis: Inhibition of Breast Cancer Progression and Metastasis by Anti‐Periostin Antibody in a Murine Model,” International Journal of Molecular Medicine 28, no. 2 (2011): 181–186. [DOI] [PubMed] [Google Scholar]
45. Nakazawa Y., Taniyama Y., Sanada F., et al., “Periostin Blockade Overcomes Chemoresistance via Restricting the Expansion of Mesenchymal Tumor Subpopulations in Breast Cancer,” Scientific Reports 8, no. 1 (2018): 4013. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1: Semiquantitative evaluation of vascular markers.

Figure S2: Digital validation of semiquantitative vascular grading.

Figure S3: Digital validation of semiquantitative periostin and FAP scores.

Figure S4: Immunofluorescence staining of CAF‐derived VEGFA and CCL2.

Table S1: List of primary antibodies.

Table S2: Gene set enrichment analysis (GSEA) of Hallmark pathways across distinct tumor components in FS (FS‐DFSP), DFSP (FS‐DFSP), and DFSP (pure‐DFSP).

Table S3: Differentially expressed genes among FS (FS‐DFSP), DFSP (FS‐DFSP), and DFSP (pure‐DFSP).

CAS-9999-0-s001.pdf^{(696.8KB, pdf)}

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

[cas70558-bib-0001] 1. WHO Classification of Tumours Editorial Board , Soft Tissue and Bone Tumours, 5th ed. (International Agency for Research on Cancer, 2020). [Google Scholar]

[cas70558-bib-0002] 2. Simon M. P., Pedeutour F., Sirvent N., et al., “Deregulation of the Platelet‐Derived Growth Factor B‐Chain Gene via Fusion With Collagen Gene COL1A1 in Dermatofibrosarcoma Protuberans and Giant‐Cell Fibroblastoma,” Nature Genetics 15, no. 1 (1997): 95–98. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0003] 3. Sirvent N., Maire G., and Pedeutour F., “Genetics of Dermatofibrosarcoma Protuberans Family of Tumors: From Ring Chromosomes to Tyrosine Kinase Inhibitor Treatment,” Genes, Chromosomes & Cancer 37, no. 1 (2003): 1–19. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0004] 4. Takahira T., Oda Y., Tamiya S., et al., “Detection of COL1A1‐PDGFB Fusion Transcripts and PDGFB/PDGFRB mRNA Expression in Dermatofibrosarcoma Protuberans,” Modern Pathology 20, no. 6 (2007): 668–675. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0005] 5. Patel K. U., Szabo S. S., Hernandez V. S., et al., “Dermatofibrosarcoma Protuberans COL1A1‐PDGFB Fusion Is Identified in Virtually All Dermatofibrosarcoma Protuberans Cases When Investigated by Newly Developed Multiplex Reverse Transcription Polymerase Chain Reaction and Fluorescence in Situ Hybridization Assays,” Human Pathology 39, no. 2 (2008): 184–193. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0006] 6. Iwasaki T., Yamamoto H., and Oda Y., “Current Update on the Molecular Biology of Cutaneous Sarcoma: Dermatofibrosarcoma Protuberans,” Current Treatment Options in Oncology 20, no. 4 (2019): 29. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0007] 7. Mentzel T., Beham A., Katenkamp D., Dei Tos A. P., and Fletcher C. D., “Fibrosarcomatous (“High‐Grade”) Dermatofibrosarcoma Protuberans: Clinicopathologic and Immunohistochemical Study of a Series of 41 Cases With Emphasis on Prognostic Significance,” American Journal of Surgical Pathology 22, no. 5 (1998): 576–587. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0008] 8. Goldblum J. R., Reith J. D., and Weiss S. W., “Sarcomas Arising in Dermatofibrosarcoma Protuberans: A Reappraisal of Biologic Behavior in Eighteen Cases Treated by Wide Local Excision With Extended Clinical Follow Up,” American Journal of Surgical Pathology 24, no. 8 (2000): 1125–1130. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0009] 9. Bowne W. B., Antonescu C. R., Leung D. H., et al., “Dermatofibrosarcoma Protuberans: A Clinicopathologic Analysis of Patients Treated and Followed at a Single Institution,” Cancer 88, no. 12 (2000): 2711–2720. [PubMed] [Google Scholar]

[cas70558-bib-0010] 10. Abbott J. J., Oliveira A. M., and Nascimento A. G., “The Prognostic Significance of Fibrosarcomatous Transformation in Dermatofibrosarcoma Protuberans,” American Journal of Surgical Pathology 30, no. 4 (2006): 436–443. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0011] 11. Hanahan D. and Weinberg R. A., “Hallmarks of Cancer: The Next Generation,” Cell 144, no. 5 (2011): 646–674. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0012] 12. Quail D. F. and Joyce J. A., “Microenvironmental Regulation of Tumor Progression and Metastasis,” Nature Medicine 19, no. 11 (2013): 1423–1437. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0013] 13. Peng R., Li Y., Gao Y., Chen D., Li Z., and Zhao Y., “Cellular and Molecular Landscape of Primary Dermatofibrosarcoma Protuberans: Insights From Single‐Cell RNA Sequencing Analysis,” Experimental Dermatology 33, no. 8 (2024): e15121. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0014] 14. Linn S. C., West R. B., Pollack J. R., et al., “Gene Expression Patterns and Gene Copy Number Changes in Dermatofibrosarcoma Protuberans,” American Journal of Pathology 163, no. 6 (2003): 2383–2395. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0015] 15. Zou X., Tang X. Y., Qu Z. Y., et al., “Targeting the PDGF/PDGFR Signaling Pathway for Cancer Therapy: A Review,” International Journal of Biological Macromolecules 202 (2022): 539–557. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0016] 16. Saijo Y., Furumoto H., Yoshida K., Nishimura M., and Irahara M., “Clinical Significance of Vascular Endothelial Growth Factor Expression and Microvessel Density in Invasive Cervical Cancer,” Journal of Medical Investigation 62, no. 3–4 (2015): 154–160. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0017] 17. Usuda K., Iwai S., Funasaki A., et al., “Expression and Prognostic Impact of VEGF, CD31 and alphaSMA in Resected Primary Lung Cancers,” Anticancer Research 38, no. 7 (2018): 4057–4063. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0018] 18. Bankhead P., Loughrey M. B., Fernandez J. A., et al., “QuPath: Open Source Software for Digital Pathology Image Analysis,” Scientific Reports 7, no. 1 (2017): 16878. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0019] 19. Subramanian A., Tamayo P., Mootha V. K., et al., “Gene Set Enrichment Analysis: A Knowledge‐Based Approach for Interpreting Genome‐Wide Expression Profiles,” Proceedings of the National Academy of Sciences of the United States of America 102, no. 43 (2005): 15545–15550. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0020] 20. Ritchie M. E., Phipson B., Wu D., et al., “Limma Powers Differential Expression Analyses for RNA‐Sequencing and Microarray Studies,” Nucleic Acids Research 43, no. 7 (2015): e47. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0021] 21. Robinson M. D., McCarthy D. J., and Smyth G. K., “edgeR: A Bioconductor Package for Differential Expression Analysis of Digital Gene Expression Data,” Bioinformatics 26, no. 1 (2010): 139–140. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0022] 22. Sonoda H., Iwasaki T., Ishihara S., Mori T., Nakashima Y., and Oda Y., “Impact of Tertiary Lymphoid Structure on Prognosis and Tumor Microenvironment in Undifferentiated Pleomorphic Sarcoma,” Cancer Science 116, no. 5 (2025): 1464–1473. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0023] 23. Schirizzi A., Arshadi A., Tolomeo D., et al., “VEGFA Status as a Predictive Marker of Therapy Outcome in Metastatic Gastric Cancer Patients Following Ramucirumab‐Based Treatment,” Biomedicines 11, no. 10 (2023): 2721. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0024] 24. Eckstein M., Epple E., Jung R., et al., “CCL2 Expression in Tumor Cells and Tumor‐Infiltrating Immune Cells Shows Divergent Prognostic Potential for Bladder Cancer Patients Depending on Lymph Node Stage,” Cancers (Basel) 12, no. 5 (2020): 1253. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0025] 25. K. S. McCarty, Jr. , Miller L. S., Cox E. B., Konrath J., and K. S. McCarty, Sr. , “Estrogen Receptor Analyses. Correlation of Biochemical and Immunohistochemical Methods Using Monoclonal Antireceptor Antibodies,” Archives of Pathology & Laboratory Medicine 109, no. 8 (1985): 716–721. [PubMed] [Google Scholar]

[cas70558-bib-0026] 26. Wang G., Zhang H., Shen X., Jin W., Wang X., and Zhou Z., “Characterization of Cancer‐Associated Fibroblasts (CAFs) and Development of a CAF‐Based Risk Model for Triple‐Negative Breast Cancer,” Cancer Cell International 23, no. 1 (2023): 294. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0027] 27. Tchou J., Zhang P. J., Bi Y., et al., “Fibroblast Activation Protein Expression by Stromal Cells and Tumor‐Associated Macrophages in Human Breast Cancer,” Human Pathology 44, no. 11 (2013): 2549–2557. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0028] 28. Jeon Y., Lee G., Jeong B. K., et al., “Proteomic Analysis Identifies Association of Periostin, a Matricellular Protein, With High Tumor Stroma and Immune Exclusion in Triple‐Negative Breast Cancer,” Cancer Research and Treatment 58, no. 3 (2026): 748–769. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0029] 29. Perea F., Bernal M., Sanchez‐Palencia A., et al., “The Absence of HLA Class I Expression in Non‐Small Cell Lung Cancer Correlates With the Tumor Tissue Structure and the Pattern of T Cell Infiltration,” International Journal of Cancer 140, no. 4 (2017): 888–899. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0030] 30. Weidner N., “Tumor Angiogenesis: Review of Current Applications in Tumor Prognostication,” Seminars in Diagnostic Pathology 10, no. 4 (1993): 302–313. [PubMed] [Google Scholar]

[cas70558-bib-0031] 31. Lugano R., Ramachandran M., and Dimberg A., “Tumor Angiogenesis: Causes, Consequences, Challenges and Opportunities,” Cellular and Molecular Life Sciences 77, no. 9 (2020): 1745–1770. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0032] 32. Ferrara N., “VEGF as a Therapeutic Target in Cancer,” Oncology 69, no. Suppl 3 (2005): 11–16. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0033] 33. Apte R. S., Chen D. S., and Ferrara N., “VEGF in Signaling and Disease: Beyond Discovery and Development,” Cell 176, no. 6 (2019): 1248–1264. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0034] 34. Zhang L., Zhang Y., Li X., Gao H., Chen X., and Li P., “CircRNA‐miRNA‐VEGFA: An Important Pathway to Regulate Cancer Pathogenesis,” Frontiers in Pharmacology 14 (2023): 1049742. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0035] 35. Qian B. Z. and Pollard J. W., “Macrophage Diversity Enhances Tumor Progression and Metastasis,” Cell 141, no. 1 (2010): 39–51. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0036] 36. Xu M., Wang Y., Xia R., Wei Y., and Wei X., “Role of the CCL2‐CCR2 Signalling Axis in Cancer: Mechanisms and Therapeutic Targeting,” Cell Proliferation 54, no. 10 (2021): e13115. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0037] 37. Ren Z., Xu Z., Chang X., Liu J., and Xiao W., “STC1 Competitively Binding betaPIX Enhances Melanoma Progression via YAP Nuclear Translocation and M2 Macrophage Recruitment Through the YAP/CCL2/VEGFA/AKT Feedback Loop,” Pharmacological Research 204 (2024): 107218. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0038] 38. Teicher B. A. and Fricker S. P., “CXCL12 (SDF‐1)/CXCR4 Pathway in Cancer,” Clinical Cancer Research 16, no. 11 (2010): 2927–2931. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0039] 39. Liang Z., Brooks J., Willard M., et al., “CXCR4/CXCL12 Axis Promotes VEGF‐Mediated Tumor Angiogenesis Through Akt Signaling Pathway,” Biochemical and Biophysical Research Communications 359, no. 3 (2007): 716–722. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0040] 40. Beider K., Bitner H., Leiba M., et al., “Multiple Myeloma Cells Recruit Tumor‐Supportive Macrophages Through the CXCR4/CXCL12 Axis and Promote Their Polarization Toward the M2 Phenotype,” Oncotarget 5, no. 22 (2014): 11283–11296. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0041] 41. Sahai E., Astsaturov I., Cukierman E., et al., “A Framework for Advancing Our Understanding of Cancer‐Associated Fibroblasts,” Nature Reviews. Cancer 20, no. 3 (2020): 174–186. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0042] 42. Liang T., Tao T., Wu K., et al., “Cancer‐Associated Fibroblast‐Induced Remodeling of Tumor Microenvironment in Recurrent Bladder Cancer,” Advanced Science (Weinheim, Baden‐Wurttemberg, Germany) 10, no. 31 (2023): e2303230. [DOI] [PMC free article] [PubMed] [Google Scholar]

[cas70558-bib-0043] 43. Kalluri R., “The Biology and Function of Fibroblasts in Cancer,” Nature Reviews. Cancer 16, no. 9 (2016): 582–598. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0044] 44. Kyutoku M., Taniyama Y., Katsuragi N., et al., “Role of Periostin in Cancer Progression and Metastasis: Inhibition of Breast Cancer Progression and Metastasis by Anti‐Periostin Antibody in a Murine Model,” International Journal of Molecular Medicine 28, no. 2 (2011): 181–186. [DOI] [PubMed] [Google Scholar]

[cas70558-bib-0045] 45. Nakazawa Y., Taniyama Y., Sanada F., et al., “Periostin Blockade Overcomes Chemoresistance via Restricting the Expansion of Mesenchymal Tumor Subpopulations in Breast Cancer,” Scientific Reports 8, no. 1 (2018): 4013. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Angiogenic and Stromal Remodeling Associated With Aggressive Fibrosarcomatous Dermatofibrosarcoma Protuberans

Sakura Shiraishi

Takeshi Iwasaki

Ken Takigawa

Kengo Kawaguchi

Hiroki Sonoda

Kohta Miyawaki

Takumi Tomonaga

Taro Mori

Makoto Endo

Koji Kato

Yasuharu Nakashima

Yoshinao Oda

ABSTRACT

Abbreviations

1. Introduction

2. Materials and Methods

2.1. Patients and Tumor Selection

TABLE 1.

2.2. Evaluation of Intratumoral Microvessels

2.3. RNA Extraction

2.4. Gene Expression Profiling and Bioinformatics Analysis

2.5. Immunohistochemistry

2.6. Immunofluorescence Staining

2.7. Statistical Analysis

3. Results

3.1. Clinicopathological Characteristics

3.2. Evaluation of Intratumoral Vasculature and Angiogenesis in FS‐DFSP

FIGURE 1.

3.3. Evaluation of Cancer‐Associated Fibroblasts in FS‐DFSP

FIGURE 2.

3.4. mRNA Expression Analysis

FIGURE 3.

3.5. Upregulation of Angiogenesis‐Related Genes in FS‐DFSP

FIGURE 4.

3.6. Evaluation of Macrophage Infiltration in FS‐DFSP

FIGURE 5.

3.7. Expression of VEGFA and CCL2 in Periostin‐Positive CAFs

4. Discussion

Author Contributions

Funding

Ethics Statement

Consent

Conflicts of Interest

Supporting information

Acknowledgments

Data Availability Statement

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases