Patient Derived Explants from Canine Mammary Gland Cancer: Histological and Immunohistochemical Evidence Using Platelet Rich Plasma as Substrate

1. Introduction

Biomedical research applied to oncology requires urgent and necessary preclinical models that accurately replicate the native architecture and microenvironment of primary solid tumors to improve understanding of tumor biology and facilitate the development of improved therapeutic approaches [1]. The use of in vitro models with cultures of immortalized human cell lines, for the study of new therapies in vitro or in vivo, is convenient but also one of the main reasons for the high failure of new drugs entering clinical trials. To move to more clinically relevant model systems, researchers have been adopting patient–derived approaches such as organoids [1,2,3] and patient-derived xenografts (PDX) [1,4,5,6].

However, several technical issues, e.g. cultivation times and high costs, limit the use of these models [7]. In human medicine, several studies recognize the potential of tissues cultured ex vivo [7,8,9,10], but in veterinary medicine it is under-explored. Patient–derived explants (PDEs) are ex vivo tumor models created by directly implanting fresh tumor tissue from patients into a supportive matrix or culture system, preserving the original tumor microenvironment. Unlike traditional 2D cultures or patient–derived xenografts (PDXs), PDEs retain the heterogeneity, architecture, and stromal interactions of the original tumor, making them highly relevant for drug testing and personalized medicine. PDEs allow for short-term screening of therapeutic responses, including immunotherapies, by maintaining immune cell populations and extracellular matrix components, which are often lost in other models. Studies have demonstrated that PDEs can predict patient responses to chemotherapy, targeted agents, and immunotherapies more accurately than cell–line–derived models [1,11]. Moreover, PDEs are being utilized to identify biomarkers for treatment stratification, particularly in aggressive cancers such as colorectal cancers [12]. However, challenges are still, including maintaining viable tissue for extended periods, ensuring reproducibility, and standardizing culture conditions. Although PDEs have a long history, general recognition of the potential of ex vivo cultured tissues to increase the clinical relevance of laboratory research [8,13,14,15,16], the PDE method has not been widely adopted to study solid tumors. Human breast tissue explant cultures provide important information for the study of breast cancer structure and phenotype, as they include the context of the surrounding microenvironment, allowing for the complete exploration of patient heterogeneity. However, the main limitation of currently available techniques remains the short–term viability of the tissue due to the loss of structural integrity [17].

PRP is a blood-derived plasma product that is easily obtained from peripheral blood samples and contains a platelet concentration approximately three to five times higher than that of whole blood. This provides a wide range of highly concentrated bioactive growth factors [18,19,20], including transforming growth factor (TGF-ß1 and TGF-ß2), platelet-derived growth factor (PDGF-AA, -AB, and -BB), vascular endothelial growth factor (VEGF-A and VEGF-C), insulin-like growth factor (IGF-1), and epidermal growth factor (EGF)These. which promote local angiogenesis, stem cell homing, cell migration, proliferation, and differentiation, along with the deposition of proteins such as collagen, which are essential for restoring normal tissue structure and function [21,22,23]. Due to the high content of growth factors in platelets, PRP has been shown to enhance cell growthoftensurpassing that of standard cultures supplemented with Fetal Calf Serum (FCS) [24]. In this study, preliminary trials used only PRP as a culture medium, with the addition of calcium gluconate to initiate the coagulation cascade and trigger growth factor release. Calcium gluconate gives PRP a gel-like consistency. After the gel retracts, PRP clot (CPRP) and PRP serum (SPRP) are obtained [25], with the latter used as a supplement to the remaining PRP.

Several studies have shown that PRP contributes to reducing regeneration times and improving tissue quality [20,26,27,28] and have promoted its wide use in clinical practice [29]. Lucarelli et al. [30] investigated the ex vivo influence of 1% up to 10% PRP as platelet gel on bone marrow stromal stem cells, showing a dose-dependent effect of PRP on cell proliferation.

Canine mammary cancer is considered by many researchers to be an excellent translational model due to the numerous similarities with breast cancer, although many aspects are still to be investigated and deepened. Canine mammary cancer is one of the main cancers diagnosed in whole. This also represents an excellent translational model of breast cancer, showing similarities both in the risk factors of onset and in terms of progression of breast cancer disease, which can be the cause of death in both women and dogs. The study of tumor in animal models allows to deepen the knowledge on prognostic factors that could also represent possible therapeutic targets both in human medicine and veterinary.

This study demonstrates that ex vivo culture derived from cancer samples obtained during surgery provides a low-cost, high-throughput model that maintains native tissue architecture, microenvironment, cell viability, and major oncogenic factors, all enhanced using PRP as a culture medium, even at low production cost, but high potential.

2. Materials and Methods

Blood was collected, using a blood bag (Terumo Blood Bags, CPDA-1) containing an anticoagulant solution (CPDA-1), from the jugular vein of a client-owned horse as part of a routine therapeutic procedure for the preparation of autologous PRP intended for the treatment of endometritis. No blood samples were collected specifically for research purposes. Residual PRP remaining after completion of the clinical treatment was subsequently used for the in vitro study. After the first centrifugation (1000–1200 rpm for 10 min, temperature at 4–10 °C), three layers were evident: the lower layer composed of red blood cells, the middle layer composed of platelets and white blood cells (buffy coat) and the upper layer of plasma. Fifty percent of the supernatant was discarded up to the buffy coat and, to obtain a platelet-enriched plasma fraction, a second centrifugation was performed. For this purpose, the plasma obtained after the first centrifugation was aspirated using a sterile syringe and transferred into 50 mL Falcon tubes. One tube was set aside to obtain fibrin–rich plasma (PRF). After the second centrifugation (2500–3000 rpm for 10 min, temperature at 4–10 °C), the upper fraction corresponded to platelet-poor plasma (PPP), whereas the lower fraction represented platelet-rich plasma (PRP). PPP was removed, while previously obtained autologous thrombin was added to PRP, thus activating the coagulation cascade. To obtain autologous thrombin, a PRP tube was isolated in which calcium gluconate (10%, Braun) was added. The mixture was incubated for 30–40 min at 37 °C and the PRP clot (CPRP) and PRP serum (SPRP) were obtained. The serum was filtered to remove cells and fibrin. The latter was used as a supplement to the remaining part of PRP obtained, as it contains significant thrombin activity.

Two canine mammary gland tumors, from patients who underwent total or partial mastectomy, were transported on ice to the laboratory of the Unit of Pathology, of the Department of Veterinary Sciences, University of Messina, within 15 min and cut into 5 x 5 x 2 mm contiguous pieces. From the first tumor, a sample was placed directly in 10% formalin (T0) while an additional four (4) were cultured in 0.9% saline solution (control group, SS) and equine heterologous PRP (PS group) for 3 and 7 days, ensuring that the tissue fragments were completely submerged in the culture medium. After each incubation period, a gross examination was performed to assess changes in volume, shape, and color, and samples were fixed in formalin for histological analysis. Hematoxylin and Eosin stain (H&E) was performed in each specimen and immunohistochemistry was carried out for cytokeratin panCK, AE1/AE3, Vimentin, PCNA. From the second tumor a sample was placed directly in 10% formalin (T0) while an additional six (6) were cultured in 0.9% saline solution (control group, SS) and equine heterologous PRP (PS group) for 7, 14 and 21 days. In this second setting, after a preliminary analysis with H&E, immunohistochemistry was carried out for Cytokeratin, Vimentin, β-catenin, , HIF–1 alpha, Ki-67, PCNA Epo (erythropoietin), EpoR (erythropoietin receptor), P63(Table 1).

Briefly, for immunohistochemical staining, samples were fixed in formalin and embedded in paraffin. Sections of 5 µm thickness on poly-L-lysine coated slides underwent microwave oven antigen unmasking in citrate buffer at pH 6 or EDTA buffer at pH 8. Hydrogen peroxide to block the endogenous peroxidases and a blocking substrate (ChemCruz—UltraCruz Blocking Reagent, Santa Cruz Biotechnology)) to block non-specific protein reactions were applied over. The sections were then incubated with the selected primary antibodies (as reported in Table 1), overnight at 4 °C. As revelation method, were then applied biotin- or HRP-conjugated secondary antibodies followed by an incubation with acetylated Streptavidin (Biospa antibiotic products, Milan; for biotin-conjugated antibody only). Diaminobenzidine (DAB) enzyme substrate, followed by nuclear counterstain with Haematoxylin, were used to stain the immune reaction.

Each round of IHC included negative reagent controls and internal controls. Each antibody was tested under different conditions as probing with and without AR, detection systems with biotin and HRP conjugated secondary Abs. The use of multiple antibodies of the same isotype and similar concentrations represented a set of irrelevant reagent controls. Immunohistochemical positivity was assessed as membranous, cytoplasmic or nuclear.

3. Results

H&E stain revealed that the first sample was a grade 1 tubular carcinoma (T0) (Figure 1).

Cultured canine mammary carcinoma explants exhibited markedly superior preservation in all incubation intervals when supplemented with PRP, as evidenced by the sustained bright pink coloration observed macroscopically for up to 21 days. In contrast, specimens maintained in physiological saline displayed a progressive shift toward a yellow hue after approximately 5 days, indicative of advancing autolytic degeneration until showing signs of putrefaction at 21 days. These macroscopic observations were corroborated by histopathological evaluation using H&E staining, which confirmed improved maintenance of tissue architecture and reduced autolytic artifacts in the PRP cultured samples.

Cytokeratin (AE1/AE3) immunolabeling demonstrated a coherent expression within luminal epithelial cell populations, showing increased immunoreactivity and intensity in PRP cultured fragments (P3–P7) compared with uncultured baseline tissue (T0). This suggests enhanced preservation of epithelial lineage markers and cellular differentiation status under PRP enriched conditions

Vimentin immunostaining highlighted consistent mesenchymal labeling across both fresh (T0) and cultured (T3–T7) samples, showing sustained stromal viability, fibroblastic cytoskeletal integrity, and preservation of epithelial–mesenchymal interactions (Figure 2). Histomorphological evaluation further revealed that PDEs maintained on gelatin scaffolds for up to 7 days retained reproducible tissue polarity, cohesive tumor cell nests, closely mirroring their uncultured counterparts. Multifocal hypercellular areas were suggestive that a proliferation had occurred, and this was further proven by the PCNA staining (Figure 3).

In the second experimental setting, H&E revealed a grade 3 tubular carcinoma (T0). At T7, T14, and T21, only PS showed marked preservation of epithelial and stromal architecture along with sustained proliferative activity. Numerous mitotic figures were consistently observed, mainly at the tumor invasion front, defined in patient-derived explant cultures as the preserved tumor–stroma interface originally present in vivo, where malignant epithelial cells are in direct contact with and infiltrate the surrounding stromal compartment. These morphological features were kept throughout the culture period up to 21 days (Figure 4).

Tumor cells remained embedded within a structurally recognizable stromal matrix, showing retention of the native tumor microenvironment in the explant model. In contrast, PS exhibited a change in the histological pattern, having more malignant characteristics and appearing with a solid appearance and increased infiltration of the surrounding stroma.

Immunohistochemical evaluation of hypoxia-inducible factor-1α (HIF-1α), Erythropoietin (Epo) and Erythropoietin receptor (EpoR) demonstrated persistent nuclear positivity in PRP samples up to T21. HIF-1α expression was mainly detected in tumor cells localized at the invasion front and in adjacent stromal cells, including spindle-shaped cells and residual immune cells. A comparable immunoreactivity was detected for Epo and EpoR, mirroring HIF-1α expression. This co-localization is compatible with activation of hypoxia-driven signaling pathways and indicates preservation of HIF-1α–dependent transcriptional activity within the patient-derived explant microenvironment.

Epithelial integrity and viability were further confirmed by cytokeratin (CK), p63 and Beta catenin immunostaining. CK was strongly expressed in cell forming tubuli and in cells multifocally distributed in the neoplastic solid growth. Beta catenin expression was continuous and homogeneous highlighting the cell boundaries in a few cells. In most neoplastic cells showed a granular, patchy and discontinuous expression along the cytoplasmic membrane, similarly to what observed in the T0 sample (Figure 5). P63 was strongly expressed in the nuclei of numerous epithelial cells, consistent with its role as a marker of epithelial stem and myoepithelial cells Ki-67 immunoreactivity was mainly located at the invasion front, where several mitoses were detectable (insets), and supported the presence of ongoing proliferative activity, in line with its known association with multiple phases of the cell cycle (Figure 6).

4. Discussion

Ex vivo cultures derived from surgically excised spontaneous canine mammary tumors is a robust, reproducible, and highly translational model for cancer research. By preserving native tissue architecture and tumor microenvironment, patient-derived explant (PDE) cultures provide a valuable alternative to conventional in vitro and in vivo models, supporting the principles of Replacement, Reduction, and Refinement (3Rs) [31]. In this context, PDEs offer a cost-effective, ethically sound platform to investigate tumor biology, drug responses, and microenvironment-dependent mechanisms with high predictive value. The development and appropriate fine-tuning of methods for preservation of ex vivo tumor cultures using PRP, or other blood components could also represent an economic model that potentially leads to the development and filing of patents. The ex vivo model is also called the "organotypic model" because it maintains the three-dimensional structure of the organ and is a model that combines the advantages of in vitro and in vivo cultures and allows the biology of cancer to be studied on tissue in its entire composition [32]. The aim is also to investigate immunoediting and the long-term effects of drugs in a set-up that closely resembles the original tumor microenvironment. It has been demonstrated that the interaction of epithelial and stromal tissues is important for epithelial maintenance and plays an important role in malignant transition to cancer [17].

In patient-derived explant cultures, the tumor invasion front corresponds to the preserved tumor–stroma interface originally present in vivo, where malignant epithelial cells remain in direct contact with, and infiltrate into, the surrounding stromal compartment. This region retains key structural, cellular, and functional features of tumor invasion, including active proliferation, stromal interaction, and spatially confined signaling responses. The persistence of a defined invasion front in patient-derived explant cultures underscores the ability of this model to faithfully preserve the dynamic tumor–stroma interface that governs invasion and disease progression in vivo, supporting its relevance for studying microenvironment-dependent tumor behavior and therapeutic responses.

In the present study, we developed and optimized an original histoculture protocol for canine mammary tumors using heterologous (equine) platelet-rich plasma (PRP) as the culture medium. Gel-like PRP matrix provided mechanical support comparable to surrounding tissue in vivo, maintaining tissue compactness while allowing diffusion of nutrients and soluble factors. This structural support was associated with sustained tissue viability and retained proliferative activity for up to three weeks, indicating the establishment of a stable and long-lasting organotypic culture.

The prolonged preservation of tissue integrity observed in our model is likely attributable to the combined effects of PRP supplementation and low-temperature storage. PRP provides a concentrated source of growth factors and bioactive molecules that support cell survival, proliferation, and tissue repair, while hypothermic conditions reduce metabolic demand, enzymatic activity, and stress-induced damage. The synergy between these two factors likely contributed to limiting apoptotic and necrotic pathways, thereby preserving tissue architecture and microenvironmental complexity over time.

The invasion front of tumor explants was characterized by the spatial convergence of proliferative, hypoxia-related, and adhesion-associated markers. Ki-67 immunoreactivity was predominantly localized at the tumor–stroma interface, consistent with active cell cycling and increased mitotic activity in this region, as previously described for invasive tumor compartments. This proliferative pattern was accompanied by sustained HIF-1α expression, supporting the presence of localized hypoxic signaling known to promote invasive growth and metabolic adaptation. In parallel, the discontinuous and patchy membranous distribution of β-catenin observed in neoplastic cells at the invasion front is compatible with altered cell–cell adhesion dynamics associated with tumor–stroma interactions and invasive behavior. Finally, the persistence of epithelial markers such as cytokeratin and p63 within this compartment indicates that invasive activity occurs in the context of preserved epithelial identity rather than complete dedifferentiation. Collectively, the coordinated expression of these markers supports the ability of the patient-derived explant model to retain biologically relevant features of the native tumor invasion front ex vivo.

Together, these findings demonstrate that the PRP-based PDE model faithfully preserves the structural, cellular, and functional features of the native tumor invasion front ex vivo. By maintaining epithelial–stromal interactions, hypoxic signaling, and proliferative activity within an intact three-dimensional context, this organotypic culture system represents a powerful platform for studying tumor progression, immunoediting, and long-term therapeutic effects. Given its simplicity, low cost, and high translational relevance, this approach holds significant potential for both veterinary and comparative oncology, as well as for future diagnostic, therapeutic, and innovation-driven applications.

5. Conclusions

Ex vivo culture of canine mammary tumors using platelet-rich plasma represents a reliable and cost-effective histoculture approach that preserves tissue architecture, cellular viability, and tumor microenvironment for extended periods. The proposed model maintains key epithelial and stromal components, supporting its suitability for studying tumor biology and microenvironmental interactions under controlled conditions. Importantly, this methodology aligns with the principles of replacement, reduction, and refinement by maximizing biological information obtained from surgically excised clinical samples.

This platform holds significant translational potential for veterinary and comparative oncology, particularly for short-term drug screening, biomarker assessment, and evaluation of tumor–stroma interactions. Further optimization and validation studies are warranted to assess predictive value for therapeutic responses and to extend culture longevity. Overall, platelet–rich plasm-based histoculture provides a promising experimental tool bridging conventional in vitro systems and in vivo models