A Novel Rat Model to Simulate Positive Margins in the Wound Bed of a Resected Sarcoma Authors: Gurtej Singh, Omer Kaymakcalan, Thomas R. Hays, Adam Abbas, Sruthi Sridevan, Jasleen Kaur, Sourish Rathi, Jocellie Marquez, Soosan Ghazizadeh, Felix Tavernier, Brendan F. Boyce, Sami U. Khan, Fazel A. Khan, Duc T. Bui, Alexander B. Dagum Published: 2024 ABSTRACT Background:One reason for local recurrence is the presence of positive surgical margins after tumor resection. An animal model accurately representing the microtumor burden will improve our understanding of these surgical margins. Using a rat model, we report a new methodology for creating microscopic tumors.Methods:Four different cell densities (1000, 10,000, 100,000, and 500,000 cells) of virus-induced Rous sarcoma XC cells (ATCC CCL-165) were topically added to 1.0 cm × 1.0 cm full-thickness wounds of male Rowett nude rats. Both Tegaderm and Tensoplast were then used to dress the wounds. After 9 days, the wound beds were excised, stained with hematoxylin and eosin, and analyzed using National Institutes of Health Image J software.Results:Better healing of the wound beds was observed for the smaller 2 cell densities (1000 and 10,000) as opposed to the higher 2 densities (100,000 and 500,000). The 2 higher cell density groups had gross identifiable tumors that extended deep through the dermis. On the other hand, the smaller cell density groups had microscopic tumor masses. Inflammation was present in all groups irrespective of the initial tumor cell densities, whereas hemorrhage was present only in the 2 higher cell density groups.Conclusions:This methodology can create a clinically relevant scenario of positive surgical margins after tumor resection. This induction method is simple, reasonably quick to use, and requires minimal surgical expertise. This approach could also develop microscopically positive margins for a much more comprehensive array of cancers. Takeaways Question: How can we develop a method to induce microscopically positive surgical margins in a rat sarcoma model to advance our understanding of tumor margin dynamics and enhance surgical precision in cancer treatment? Findings: Introducing tumor cells at a density of 10,000 cells into full-thickness skin wounds in a rat model can result in the formation of small, isolated tumor masses, effectively simulating microscopically positive margins. Meaning: Our methodology offers a versatile template for cultivating various tumor types, thereby facilitating the development of effective cancer therapies. INTRODUCTION Soft-tissue sarcoma (STS) is a cancer of mesenchymal cell origin that makes up 1% of all solid neoplasms in adults and 7% in children. It has an estimated annual incidence of 5 cases per 100,000 people.1 In 2017, there were more than 12,000 new cases in the United States and nearly 5000 deaths, with a reported 5-year survival rate of 64%.2 The American Cancer Society estimates a continued rise in STS incidence and mortality over time.3 STS is notorious for its high rate of recurrence and lung metastasis.1 The principal treatment, which has remained relatively unchanged for decades, consists of complete macroscopic surgical resection with negative pathological margins of at least 1 mm. This approach, known as recurrence 0 (R0), ensures the absence of gross or microscopic tumor in the primary tumor bed and may be followed by adjuvant radiation or chemotherapy.1,4 Multiple studies have identified positive surgical margins to be the most crucial predictor of local recurrence. However, the optimal distance between sarcoma and the nearest negative margin remains debated.2 Although it is every surgeon’s goal to complete tumor resection with defined negative margins, an unfortunate 30% of patients with resected tumors develop recurrence due to positive margins.5 In a prospective cohort study, 23% of patients were found to have positive margins after surgical resection of extremity STS, among which 26% went on to develop local recurrence.6 Positive margins may also arise when a mass is removed without prior biopsy, only to later be identified as STS. In instances when the tumor involves a critical structure like a major blood vessel or nerve, positive margins may be unavoidable and adjuvant therapy is given to limit major morbidity and loss of function.7 A retrospective review of patients with local STS recurrence reported a more than 10% lower 5-year survival rate for patients with positive margins compared with negative margins (P = 0.01).8 Positive margins profoundly affect mortality, making it essential to create an accurate animal model for study. A tumor induction method that mimics the microtumor burden in surgical wound beds with microscopically positive (R1) margins would be highly valuable. This would enable the development of effective therapies to improve patient care. Although animal cancer research is most commonly performed with mice, rats are still relatively inexpensive and considerably cheaper than some alternatives such as dogs, pigs, or primates.9 Furthermore, their larger size and clinically more relevant physiology and toxin excretion pathways make rats more appropriate for surgical or drug-related studies of human tumors than mice.10 Currently, there is no published method detailing how to induce microscopically positive surgical margins in a rat sarcoma model. Many previous studies have simulated positive surgical margins by partially resecting an established sarcoma. Other techniques have sought to induce sarcoma via transgenic mice with STS-producing mutations or primary tumor induction via injection of carcinogens; however, these models have only ever simulated grossly positive margins (R2).11,12 In this article, we describe a novel method to induce microscopic tumors derived from the XC sarcoma cell line in the surgical wound bed of a nude rat. Our model simulates the clinically relevant scenario of microscopically positive surgical margins after tumor resection and may serve as an example for future studies to develop effective cancer therapies for patients with positive surgical margins. MATERIALS AND METHODS Cell Culture Virus-induced Rous sarcoma XC cells (ATCC CCL-165) were obtained from American Type Culture Collection and cultured and maintained in Eagle’s Minimum Essential Medium with 10% fetal bovine serum. The cultures were maintained in T-75 flasks under standard conditions. Rat Models and Creation of Tumor Wound Beds Male Rowett nude (RNU) rats, weighing between 200 and 300 g, were used for this study. All animal procedures received approval from the Stony Brook University Animal Care and Use Committee, following Institutional Animal Care and Use Committee protocol (1020829). This protocol addresses cancer recurrence and the use of vacuum-assisted closure therapy in a rat model. The experiments were conducted adhering to the ethical standards outlined in the Animal Research: Reporting of In Vivo Experiments guidelines.13 The rats were housed under standard conditions with ad libitum access to food and water. One to four full-thickness wounds (1.0 cm × 1.0 cm), penetrated through the panniculus carnosus, extending down to the investing fascia of the back muscles served as the base of the wound. We utilized a total of 9 rats, with each rat undergoing different treatments across 4 wound sites per rat. Each wound site received a different concentration of tumor cells to evaluate the effects of varying cell densities on tumor development. Specifically, the distribution was as follows: Rats 1–3: Each rat had 4 wounds, with each wound receiving one of the following cell densities: 1000; 10,000; 100,000; and 500,000 cells. Rats 1–3: Each rat had 4 wounds, with each wound receiving one of the following cell densities: 1000; 10,000; 100,000; and 500,000 cells. Additional configurations: Other rats were used to test different configurations and ensure robustness of data across the same cell densities:-Rat 4 with 1 wound with 500,000 cells.-Rat 5 with 2 wounds: 1 with 500,000 cells and 1 with 1000 cells.-Rat 6 with 2 wounds: each with 100,000 cells.-Rat 7 with 1 wound with 1000 cells.-Rat 8 with 1 wound with 10,000 cells.-Rat 9 with 2 wounds, each with 10,000 cells. Because both wounds had similar outcomes, only 1 was included in our results to maintain a consistent number of 5 replicates per cell density. Additional configurations: Other rats were used to test different configurations and ensure robustness of data across the same cell densities: -Rat 4 with 1 wound with 500,000 cells. Rat 4 with 1 wound with 500,000 cells. -Rat 5 with 2 wounds: 1 with 500,000 cells and 1 with 1000 cells. Rat 5 with 2 wounds: 1 with 500,000 cells and 1 with 1000 cells. -Rat 6 with 2 wounds: each with 100,000 cells. Rat 6 with 2 wounds: each with 100,000 cells. -Rat 7 with 1 wound with 1000 cells. Rat 7 with 1 wound with 1000 cells. -Rat 8 with 1 wound with 10,000 cells. Rat 8 with 1 wound with 10,000 cells. -Rat 9 with 2 wounds, each with 10,000 cells. Because both wounds had similar outcomes, only 1 was included in our results to maintain a consistent number of 5 replicates per cell density. Rat 9 with 2 wounds, each with 10,000 cells. Because both wounds had similar outcomes, only 1 was included in our results to maintain a consistent number of 5 replicates per cell density. These cells were added topically to wounds using a 200-µL pipette tip to each of these 4 groups.14–16 The volume of media used was consistent at 100 µL across all wounds to ensure uniform conditions. The diverse configurations across rats aimed to mitigate individual biological variability and enhance the reliability of our finding. The wounds were dressed in Tegaderm (3M, Minneapolis, MN) and Tensoplast (BSN Medical, Hamburg, Germany) for a period of 9 days. By day 9, the use of higher concentrations resulted in a grossly evident tumor burden, indicating that the model no longer represented a microscopic tumor scenario by that time. The dressings were changed every third day (Fig. 1). The volume of 100 µL was chosen as it consistently covered the entire wound bed without spillover onto the dermis. The Tegaderm was applied over and adhered to the skin, maintaining this barrier. The rats were euthanized on day 9, and the wound bed area was excised and fixed in 4% paraformaldehyde for 24 hours. Administration of tumor cells onto full-thickness wound bed. A, Full-thickness open skin wound (1 cm × 1 cm), before inoculation of tumor cells in 100 µL of media. B, Wound covered with Tegaderm after tumor cell inoculation. C, Picture of an RNU rat with 4 wound sites. Specimen Processing and Analysis Upon euthanasia at day 9—chosen based on ethical considerations and preliminary observations indicating substantial tumor burden—wound bed areas were excised. The decision to focus on a specific timeframe and the wound bed area was driven by our study’s objective to explore how different concentrations of cancer cells influence tumor development within these beds. We acknowledge a limitation: this approach does not examine systemic spread. The fixed excised tissue samples were bisected longitudinally, and one half was embedded in paraffin. Five-micrometer-thick sections were stained with hematoxylin and eosin and analyzed for the presence of tumor cells by 2 pathologists who were blinded to the treatments. The tumor length and depth were measured, and tumor areas were calculated by multiplying the tumor length by the depth of invasion. Data Analysis Means, SDs, medians, and the distribution of the tumor areas were calculated. We assessed other variables, including granulation tissue, ulceration, inflammation, hemorrhage, and fat tissue, and we compared these using a chi-square test or a Kruskal-Wallis test when data was nonparametric. SPSS 28 software was used to perform the analysis. RESULTS Observations on Tumor Development and Healing Dynamics In wounds with lower cell densities (1000 and 10,000), healing of the skin was observed in most cases. However, wounds treated with higher cell densities exhibited brown discoloration and incomplete healing, with tumor cells spreading beyond the wound borders. Grossly identifiable tumors, which formed in groups with initial cell densities of 100,000 cells/100 µL of media or higher, extended deep into the wound bed. These were not observed in groups treated with lower cell densities (1000 and 10,000/100 µL of media). Despite this, the depth of tumor invasion remained confined and did not surpass the fascial layer beneath the subcutaneous fat across all groups (Fig. 2). Excised wound beds from RNU rats, treated with varying densities of XC sarcoma cells, were imaged on day 9 postinoculation. A, Initial cell density of 1000 cells and 500,000 cells. B, Initial cell density of 10,000 cells. C, Initial cell density of 100,000 cells. Quantitative Analysis of Tumor Areas The analysis of tumor area values, which were nonparametric, is presented in Figure 3. To compare the tumor areas across the 4 groups, the Kruskal-Wallis test was used. The means and medians of the calculated tumor areas for all 4 groups are detailed in Table 1, revealing that the mean tumor area significantly increased with higher numbers of inoculated tumor cells (P = 0.015). Specifically, tumors were found in 20% of rats inoculated with 1000 tumor cells, 60% with 10,000 cells, and 100% with 100,000 or more cells. Table 1.Means and Median for Different Cell DensitiesInitial Cell DensityMean (mm²)Median (mm²)SD (mm²)Minimum (mm²)Maximum (mm²)Observations10000.03200.07200.16Very low variability: a majority of samples showed no tumor, indicating minimal tumor growth at this density10,0001.9360.084.010109.1Increased variability, along with the presence of a notable outlier, can most likely be attributed to the methodology used in the analysis100,0009.573410.0150.3621.08High variability with substantial tumor growth in some samples, indicating a more pronounced response at this density500,00048.555417.4932165Consistently high measurements indicating strong tumor growth across all samples Means and Median for Different Cell Densities Areas of tumors in the groups of nude rats (n = 5 replicates per cell density) (P = 0.015). Associated Pathological Findings Furthermore, the occurrence of hemorrhage followed a similar pattern; it was absent in the 1000 and 10,000 cell groups but appeared in groups with higher cell densities (100,000 and 500,000). Inflammation was observed at all wound sites irrespective of the inoculated cancer cell density. However, there was no evident correlation between the presence of granulation tissue and the density of tumor cells. Similarly, no substantial correlation was identified concerning the extent of tumor invasion into fascia across the different rat groups. Notably, tumor invasion did not extend beyond the fascial layer in any of the specimens (Table 2). Table 2.Histological Features of Tumor Wound Beds for Different Initial Cell DensitiesDensity of Virus-induced Rous Sarcoma Cells Injected100010,000100,000500,000Histological featureCount (%)Count (%)Count (%)Count (%) P Tumor identified*1 (20)3 (60)5 (100)5 (100)0.015Ulcer1 (20)0 (0)0 (0)3 (60)0.058Granulation tissue5 (100)3 (60)4 (80)3 (60)0.402Inflammation5 (100)5 (100)5 (100)5 (100)Scar4 (80)2 (40)3 (60)1 (20)0.261Hemorrhage0 (0)0 (0)1 (20)1 (20)0.528Panniculus4 (80)4 (80)3 (60)4 (80)0.849*Significant difference of frequencies over different tumor densities. Histological Features of Tumor Wound Beds for Different Initial Cell Densities Significant difference of frequencies over different tumor densities. Histological Insights Representative histological images from each experimental group are provided in Figure 4. These images illustrate the varying degrees of tumor development and associated tissue responses, offering visual confirmation of the quantitative data presented. Histological images from day 9 corresponding to different cell densities added on day 0: (A) 1000 tumor cells were added on day 0, and the image represents day 9; (B) 10,000 tumor cells were added on day 0, and the image represents day 9; (C) 100,000 tumor cells were added on day 0, and the image represents day 9; (D) 500,000 tumor cells were added on day 0, and the image represents day 9. DISCUSSION Positive surgical margins after STS resection serve as the most principal risk factor for local recurrence and adversely affect patient morbidity and mortality.2 Here, we report a novel nude rat sarcoma model in which microscopic tumor growth was induced at a surgical site. We discovered that inoculating a full-thickness skin wound with 10,000 XC sarcoma cells produced small, isolated tumor masses. In contrast, using cell densities of 100,000 or more cells resulted in the formation of large, gross tumor masses. Our novel tumor induction model offers many distinct advantages over other models and produces microscopic tumors in a relatively short 9-day period. In contrast, previous studies have reported that female RNU rats took up to 4 weeks to develop tumors. A different modification involves subcutaneously inoculating rats with small tumor fragments embedded in Matrigel, followed by 500 cGy of whole-body irradiation.17 Primary induction methods typically are associated with longer and more variable tumor latency periods. For example, 1 study involved repeated weekly injections of Wistar rats with 1 mg of 4-(hydroxyamino) quinolone 1-oxide for a period of 4 weeks. This chemical induction of undifferentiated pleomorphic sarcoma resulted in an 87% occurrence rate of tumors between 16 and 48 weeks postinduction.18 Another approach saw subcutaneously implanted materials such as polyethylene or polyvinylchloride in Fischer rats. These induced sarcomatous tumors appeared at 26 and 40 weeks postimplantation, respectively.19 Most animal studies simulate positive surgical margins by partially resecting an established sarcoma. For example, positive margins were simulated in a murine model by resection of 95% of the volume of a subcutaneous flank tumor.5 In another study, sarcomas were induced in mice by injecting an adenovirus carrying Cre-recombinase into the hindlimb, targeting oncogenic mutations in p53, BRAF, and K-RAS. After complete tumor excision, a 3 mm³ piece of the tumor was reimplanted, but these methods only simulated grossly positive margins (R2).11 Tumors may also be produced de novo from the host’s own cells by multiple methods, such as injection of chemical, viral, or physical carcinogens. Transgenic murine models also have been developed using Kras, Trp53, or Pten mutations to induce growth of various STSs.12 Although primary induction may result in metastatic disease, it is associated with a much longer and more variable tumor latency period and, therefore, may not be practical for continuous monitoring.20,21 In addition, research personnel must take extra precautions to avoid hazardous exposure while handling carcinogens. The creation of a 1 cm × 1 cm full-thickness skin wound is simple and does not require advanced surgical expertise. It also enables control of tumor location, and the superficial location of the tumor allows for easy access and continuous monitoring of tumor growth. Our induction method in nude/athymic (rnu/rnu) rats allows for enhanced growth, as the impaired adaptive immune system enhances tumor progression.22 Although mice are utilized in 95% of all animal cancer research studies, rats may serve as an alternative, with unique advantages.20 Rats’ heavily documented physiology and toxin excretion pathways closely mimic those of humans, which may be particularly useful for testing adjuvant chemotherapy.23 Other suitable animal models, such as monkeys or pigs, could be entertained, but the financial costs are a considerable limitation.9 A rat sarcoma model that accurately recapitulates positive (R1) surgical margins may serve as the foundation for future research to investigate how different therapies impact microscopic tumors. For example, negative pressure wound therapy is currently contraindicated by device manufacturers for use in malignant wounds due to concerns about local recurrence; however, clinical evidence to support this recommendation is lacking.24 Our tumor induction method could be used to simulate positive surgical margins, and negative pressure wound therapy could be applied to the wound to determine if this contraindication is valid and to better determine safety guidelines. Although STSs are rather aggressive malignancies with a high rate of local recurrence, they are rare tumors. Other cancers, such as breast and melanoma, are much more common and can also have high recurrence rates.25,26 The National Cancer Institute estimated that breast cancer and melanoma would, respectively, comprise 15.3% and 5.6% of all new cancer diagnoses in the United States in 2020.27,28 It is possible that our method could be used to create microscopically positive margins using breast, skin, and a much wider array of cancer cell lines to study local recurrence phenomena. This research is particularly relevant to both plastic and oncologic surgeons involved in postoperative management, where decisions about the timing of adjuvant treatments are critical. A model simulating microscopically positive margins would be extremely valuable, providing a detailed framework to evaluate the efficacy of various reconstructive and adjuvant therapies. Such insights could markedly refine the strategies for initiating adjuvant therapy, thereby optimizing patient outcomes and reducing the rates of tumor recurrence. Limitations Our tumor induction method and nude rat sarcoma model have some limitations. Rat sarcomas do not perfectly resemble the characteristic patterns, histology, and physiology of human sarcomas.29 This may lead to suboptimal clinical translation of any results. In addition, the use of athymic nude rats with reduced adaptive immunity may not accurately predict the clinical efficacy of cancer therapies in immunocompetent individuals or the natural growth of recurrent sarcoma. However, the nude (rnu/rnu) rat lacks only a normal thymus and functionally mature T cells and has sufficient innate immunity to be able to release acute inflammatory cytokines to potentially influence tumor growth.30,31 The acute inflammatory reaction induced by a full-thickness surgical wound may enhance tumor growth via release of transforming growth factor-β and increased COX-2 expression.5,32,33 Kottke et al34 reported that recurring tumors in rodents have vastly different phenotypes from the original tumor. Additionally, our literature search revealed no strict definitions for sizes of gross tumors versus microscopic tumor foci. Given that this is a limitation of not just our study, but also that of the field in general, future studies are needed to identify a consistent phenotypic definition of both gross tumor burden and microscopic deposits of tumor cells. Furthermore, it is possible that microscopic deposits of tumor cells formed in the groups of rats with smaller cell densities (1000 and 10,000/100 µL of media) but were not identified, because they were not present in the 5-µm sections that were cut and examined. In future studies, we plan to inoculate fewer tumor cells and use a more accurate method of tumor detection (3-dimensional volumetric analysis of tumor volume), rather than the currently used strategy (area in 5-µm-thick tissue sections). Another limitation is the absence of a formal power analysis to determine the optimal sample size for detecting meaningful differences between treatment groups. This suggests that some of the findings might need further validation through larger, more robust studies. This work may be of interest to both plastic and oncologic surgeons. Although many plastic surgeons do not participate in the excision of tumors; they are, however, an integral part of the subsequent postoperative management. It is often the responsibility of the plastic surgeon to determine the appropriate timing for initiating radiation therapy or chemotherapy, as these decisions are contingent upon the healing of the postoperative wound. A model featuring a microscopically positive margin would be particularly valuable for plastic surgeons. It would allow them to study various reconstructive approaches with and without adjuvant treatment and their impact on the incidence and rate of tumor recurrence, thereby informing the optimal timing for resuming adjuvant therapy. CONCLUSIONS This pilot study introduces a novel method for the inoculation of an XC sarcoma cell line into full-thickness skin wounds in male RNU rats. The results demonstrate the potential of varying cell densities to simulate different tumor environments: inoculating 10,000 cells into a wound results in small, isolated tumor masses that effectively simulate microscopically positive margins, whereas higher densities such as 100,000 cells or more lead to the formation of larger, gross tumor masses. However, it is crucial to note that due to the absence of a formal power analysis and the exploratory nature of this study, the results should be interpreted with caution, as they may not have been sufficiently powered to definitively show these differences. Despite these limitations, our model offers substantial insights and serves as a promising tool for future oncological research. It provides a basic framework for developing effective cancer therapies and creating additional rat models to investigate local recurrence across a broader spectrum of cancers beyond STS. These models could be particularly useful for studying the dynamics of tumor development and testing therapeutic interventions in a controlled environment. DISCLOSURES The authors have no financial interest to declare in relation to the content of this article. This project received funding from Targeted Research Opportunities Clinical Research Awards, Stony Brook School of Medicine.