In this protocol, we describe a mouse model of incomplete surgical resection of soft tissue sarcoma for testing (neo)adjuvant therapies.
Surgery is often the first treatment for many solid tumors. However, local relapses frequently occur following primary tumor resection, despite adjuvant or neo-adjuvant therapies. This occurs when surgical margins are insufficiently tumor-free, resulting in residual cancer cells. From a biological and immunological perspective, surgery is not a null event; the wound healing environment is known to induce both pro- and anti-tumorigenic pathways. As a consequence, preclinical models for drug development aimed at preventing local relapse should incorporate surgical resection when testing new (neo)adjuvant therapies, to model the clinical settings in patients treated with surgery.
Here, we describe a mouse model of incomplete surgical resection of WEHI 164 soft tissue sarcoma that allows testing of (neo)adjuvant therapies in the setting of a wound healing response. In this model, 50% or 75% of the tumor is removed, leaving behind some cancer tissue in situ to model gross residual disease after surgery in the clinical setting. This model allows testing therapies in the context of surgery while also considering the wound healing response, which may affect the efficacy of (neo)adjuvant treatments. The incomplete surgical resection results in reproducible regrowth of the tumor in all mice in the absence of adjuvant therapy. Adjuvant treatment with checkpoint blockade results in reduced tumor regrowth. This model is thus appropriate for testing therapies in the context of debulking surgery and its associated wound healing response and can be extended to other types of solid cancer.
Surgery remains the main treatment option for many solid tumors1, including soft tissue sarcoma2,3. Despite improvements in cancer surgery techniques, and combinations with (neo)adjuvant therapies, there is still a high risk of cancer relapse and metastasis following primary tumor resection4,5. In soft tissue sarcoma, relapses occur particularly locoregionally, at the site of surgery, resulting in increased morbidity and mortality. In the clinical setting, it can be difficult to obtain wide enough margins (e.g., due to anatomical constraints), resulting in incomplete resection and subsequent tumor recurrence6. Surgical stress and the subsequent process of wound healing are known to create an immunosuppressive tumor microenvironment favorable for tumor recurrence7,8. Therefore, the discovery and development of new therapies for soft tissue sarcoma, particularly immunotherapies, should ideally take the surgical wound healing response into account.
Most preclinical studies for adjuvant therapies are initially carried out using subcutaneous syngeneic or xenotransplant mouse models, without incorporating the surgical stress and wound healing response9,10. Therefore, we developed a syngeneic subcutaneous mouse soft tissue sarcoma model incorporating incomplete surgical resection. WEHI 164 fibrosarcoma cells are inoculated subcutaneously, and once tumors are established, we remove 50-75% of the tumor bulk (Figure 1A-E). Tumors consistently re-grow from the remaining tumor. This model allows for testing adjuvant therapies while considering the effect of surgical stress and wound healing. Similar surgical models of incomplete resection have been used in a number of studies by several groups and found to be reproducible and effective11,12,13. Here, we provide a detailed description of this protocol.
Animals used in these experiments were obtained from the Animal Resource Centre (Perth, Western Australia). Animals were maintained under standard pathogen-free conditions at the Harry Perkins Institute of Medical Research Bioresources North Facility (Perth, Western Australia). All experiments were carried out following the protocol as approved by the Harry Perkins Institute of Medical Research Animal Ethics Committee. BALB/c mice of 8-12 weeks of age were used in these experiments. The WEHI 164 fibrosarcoma cell line was obtained from CellBank Australia (Westmead, NSW).
1. Inoculation of cells
2. Partial surgical resection of the tumor
NOTE: This protocol requires TWO researchers; one for surgical procedures (SURGEON), and another for mouse monitoring (ASSISTANT).
Tumor growth to a size of 50 mm2 is an ideal size for partial debulk. The incomplete surgical resection of 50 mm2 tumors results in 100% (n=5) reproducible regrowth of the tumors in the absence of adjuvant immunotherapy (Figure 4A). We next used the model to test adjuvant immunotherapies using antibodies against checkpoint molecules Cytotoxic T Lymphocyte Associated Protein 4 (CTLA-4) and Programmed Death Receptor 1 (PD-1). Treatment of mice with anti-CTLA-4 or anti-PD-1 resulted in a cure rate of 80% and 25% (n=4-5 per group), respectively (Figure 4B, 4C). The response with anti-PD-1 provides an opportunity to test novel combinations to improve the response rate further.
Figure 1: Schematic diagram of partial surgical resection of the tumor. (A) BALB/c mice are inoculated with 5 x 105 WEHI-164 cells on the lower right flank. (B) When the tumor reaches 50 mm2, surgery can commence. (C) The tumor is partially resected (50 % shown). (D) The surgical site is closed with clips. (E) Adjuvant therapy can be administered, intravenously, intraperitoneally (shown) or intratumorally in the wound area. Please click here to view a larger version of this figure.
Figure 2: Representative images of the surgery set up. (A) A whole image of the surgery set up showing the surgical tools (listed in step 2.1) and the anesthetic machine. (B) A snapshot image of the surgical table showing all materials within an easy reach. (C) A heating chamber and a heating pad for mouse recovery. Please click here to view a larger version of this figure.
Figure 3: Representative pictures of partial tumor debulk technique. (A) A fully anesthetized mouse with a tumor of 50 mm2 in size before surgery. (B) Incision site 3 mm away from the tumor; 1 cm incision. (C-D) Opening of the wound by gently holding the skin on the tumor bearing side using tweezers, and "inverting" the tumor so that it is visible outside. Please click here to view a larger version of this figure.
Figure 4: Tumor regrowth following incomplete tumor resection and immunotherapy. (A) Tumor regrowth curves of partially resected WEHI-164 tumors in the absence of adjuvant immunotherapy. (B-C) Tumor regrowth after surgery and adjuvant treatment with anti-CTLA-4 (B) or anti-PD1 (C). The dotted line indicates the day of surgery. Please click here to view a larger version of this figure.
We provide a protocol for a mouse model of incomplete surgical resection of soft tissue sarcoma to test peri-operative therapies. We also standardized the surgical incision to allow assessment of wound healing between mice following treatment.
Tumor placement is an important part of this protocol. We have opted for a subcutaneous tumor model to allow easy surgical access to the tumor site and administration of local therapies with minimal burden on the mice. It is also important to ensure that the tumors grow in the subcutaneous space and not within the peritoneum, which can result in unexpected morbidity and mortality.
When choosing a tumor cell line for this protocol, we advise that the cells when grown in vivo form a solid mass (e.g., WEHI-164 model), rather than a semi-solid mass (such as the B16 model) as it is technically difficult to partially resect. In addition, if the tumor begins to grow through the skin (usually seen in tumors larger than 100 mm2), debulking is not recommended as the skin may become necrotic and not heal well after surgery. We have overcome this problem by debulking tumors once they reach 50 mm2 in size.
As our model can be used to assess the effect of wound healing on therapy, we propose a control/sham group as a comparison. The control may be unaltered tumor, or sham surgery which would only have the skin incision, exposure of the tumor, and wound closure without partial tumor debulk. This sham control group could be used when discerning the effect of surgery-induced inflammation and wound healing from the partial debulk on the treatment outcome.
For successful partial debulking surgery, some technical points need to be considered. An important aspect is the correct implantation and growth of the tumor. Tumors need to be implanted on the lower-right flank, away from the hind leg. Tumors that are implanted too close to the hind leg can interfere with their ability to walk and can result in extra force on the clips causing them to come loose. In addition, consistency in tumor size is critical in order to avoid variability in the relative percentage of debulking. We chose to perform surgery with tumors that have a size of 50 mm2, to make surgery technically straightforward, although we envisage that partial resection on smaller tumors is feasible. To prevent inconsistency in tumor size, the used cell line needs to be passaged following the appropriate standard cell culture techniques, and the researcher needs to be adequately trained in the proper tumor inoculation technique. When extending this protocol to other subcutaneous tumor models, the physical characteristics of the tumor are of importance. For example, we found that cell lines that give rise to soft, gelatinous tumors (e.g., M3-9-M rhabdomyosarcoma, and B16 melanoma15) are technically challenging to debulk.
There are also technical points that need to be considered during surgery. Mice need to be adequately anesthetized to prevent movement during the procedure. Apart from the impost that inadequately anesthetized mice will endure, any movement of mice during the procedure can make the surgical resection difficult, resulting in variability in the size of tumor removed between mice. In addition, mouse respiratory rate should be carefully monitored during surgery Isoflurane concentration should be adjusted to maintain the appropriate depth of anesthesia. Therefore, an assistant is always needed during the surgical procedure to monitor the breathing rate during surgery, and to ensure an adequate level of anesthesia. The size of the incision needs to be consistent in order to avoid variability in the wound healing response. We found that a 1-1.5 cm incision is sufficient for tumor debulking, with a minimal chance of wound dehiscence.
Our model of partial resection mimics residual disease remaining after surgery as seen in the clinical setting of many solid tumors and offers advantages over traditional syngeneic mouse models by taking into account the effect of surgical wound healing. In addition, existing traditional models of surgery have used complete tumor resection, which does not always result in tumor recurrence16. Other researchers have successfully used partial resection models using other cancer cell lines11,12,13, underscoring the robustness of this method. Furthermore, it has been demonstrated that partial resection, but not complete resection, resulted in protective anti-tumor immune memory when adjuvant therapy is given12 which was attributed to the persistence of antigens from the residual tumor.
This model is designed to study the effect of inflammation and wound healing on therapy. Our debulking approach clinically resembles clinical situations where gross residual disease is left behind after surgery (R2 resection), rather than macroscopically complete resection with microscopic residual disease (R1 resection). For example, surgical resection in invasive soft tissue sarcoma can result in positive margins when the tumor is located next to critical structures such as nerves, arteries or adjacent organs, precluding complete resection with wide margins17. Surgery models for resection resulting in microscopic positive margins have been published13; our protocol can be used to study the effect of the wound healing response on therapy when macroscopic residual disease is present.
A limitation of our model is that it does not give rise to distant relapse and micrometastasis, which is common after surgery in solid tumors such as breast cancer or pancreatic cancer. Other surgery models, such as the murine breast cancer model 4T118,19,20 or murine models of de novomammary cancer metastasis21 are better suited to investigate systemic relapse after local resection. Another limitation is that this protocol is for subcutaneous models and thus does not allow assessment of tissue-specific pathology. For this purpose, orthotopic tumor mouse models are appropriate7,22,23. However, orthotopic models are more challenging and usually involve greater impost to mice, and are more laborious and costly22. Subcutaneous models are well suited to assess the effects of (neo-)adjuvant therapies, either systemically or locally, on local cancer relapse, in a cost-effective and relatively high-throughput manner with minimal impost to the animals.
The incomplete partial resection as outlined in this protocol is useful for testing adjuvant therapies while incorporating surgical wound healing as a factor, a variable which is often overlooked.
The authors have nothing to disclose.
This work is supported by grants from the Sock it to Sarcoma! Foundation, the Australian and New Zealand Sarcoma Association, the Children's Leukemia & Cancer Research Foundation and Perpetual Philanthropy. W.J.L is supported by a Simon Lee Fellowship and a research fellowship from the National Health and Medical Research Council, and the Cancer Council WA.
26 gauge 0.5 mL insulin syringe | Becton Dickinson, Australia | 326769 | None |
2-Mercaptoethanol | Life Technologies Australia Pty Ltd | 21985023 | None |
Anaestetic gas machine | Darvall Vet, Australia | SKU: 2848 | None |
Anti-CTLA-4 | BioXcell, USA | BE0164 | None |
Anti-PD-1 | BioXcell, USA | BP0273 | None |
Buprenorphine Hydrochloride Injection, 0.3mg/mL | RB healthcare UK Limited, UK | 55175 | Prescription order |
Chlorhexidine Surgical Scrub 4% | Perigo Australia, Australia | CHL01449F(scrub | None |
Fetal Bovine serum | CellSera, Australia | AU-FBS-PG | None |
Forceps Fine 10.5 cm | Surgical house, Western Australia | CC74110 | None |
Forceps Fine 12 cm Serrated | Surgical house, Western Australia | CC74212 | None |
Forceps Halsted 14 cm | Surgical house, Western Australia | CD01114 | None |
Heating chamber | Datesand Ltd, UK | Mini-Thermacage | None |
HEPES (1M) | Life Technologies Australia Pty Ltd | 15630080 | None |
Isoflurane | Henry Schein Animal Health, Australia | SKU: 29405 | Prescription order |
Lubricating Eye Ointment | Alcon | n/a | None |
Penicillin/streptomycin 1000X | Life Technologies Australia Pty Ltd | 15140122 | None |
Phosphate Buffered Solution 10x | Life Technologies Australia Pty Ltd | 70013-032 | None |
Reflex 7mm Clips | Able scientific, Australia | AS59038 | None |
Reflex 7mm Wound Clip Applicator | Able scientific, Australia | AS59036 | None |
Reflex Wound Clip Remover | Able scientific, Australia | AS59037 | None |
Rodent Qube Anesthesia Breathing Circuit | Darvall Vet, Australia | #7885 | None |
Roswell Park Memorial Institute (RPMI) 1640 Medium + L-glutamine | Life Technologies Australia Pty Ltd | 21870092 | None |
Scissors Iris STR 11 cm | Surgical house, Western Australia | KF3211 | None |
Scissors Iris STR 9 cm | Surgical house, Western Australia | JH4209 | None |
Small Induction Chamber | Darvall Vet, Australia | SKU: 9630 | None |
TrypLE express 1x | Life Technologies Australia Pty Ltd | 12604-021 | None |