Angiographic Study and Therapeutic Embolization of Soft-Tissue Fibrosarcoma in a Dog: Case Report and Literature

Case Report

A 5-year-old, male Serra de Aires dog was referred to the Minimally Invasive Surgery Centre, Cáceres, Spain because of a 20-day history of a large mass in the left thigh with associated pain and lameness. The additional history included weight loss and progressive anorexia. Rectal temperature was 39.5°C, and heart rate was 92 beats per minute. The mucous membranes appeared pale. A moderate lameness of the left pelvic limb was identified on general physical examination, and a large mass 15 to 20 cm in diameter was palpable on the medial side of the left thigh. Local skin color and temperature were normal; however, the left femoral artery pulse was weak, and the lower part of the limb distal to the mass was cold in temperature.

Radiographs of the thorax and left hind limb were obtained. No abnormalities were found on chest radiographs. Radiographs of the left hind limb identified swelling at the proximal, medial side of the left thigh without significant calcification [Figure 1]. The left coxofemoral joint was intact. Neither bone destruction nor periosteal reaction or new bone formation were identified. Three attempts of fine-needle aspiration (FNA) biopsy were made over a 2-week period. The aspirates consisted of dark, uncoagulated blood in each procedure. Cytopathological studies were inconclusive; a surgical biopsy was offered to the owner but was declined due to financial restrictions.

Based on the authors’ investigations regarding the utility of angiography in veterinary oncology, complimentary angiography was offered and accepted by the owner of the dog. Following a 24-hour fast, the dog was premedicated with intramuscular (IM) acepromazine (0.05 mg/kg body weight), meperidine (5 mg/kg body weight), and atropine (0.04 mg/kg body weight). Hydration was maintained with intravenous (IV) normal saline. Induction of anesthesia was performed with propofol (4 mg/kg body weight, IV). After the dog was endotracheally intubated, anesthesia was maintained with isoflurane (2% to 2.5% in oxygen [O₂]); blood pressure, electrocardiogram, O₂ saturation, end-tidal carbon dioxide (CO₂), and body temperature were monitored closely throughout the procedure.

Under sterile conditions, right superficial femoral arterial access was established by percutaneous placement of a 6-Fr introducer sheath.^a Under fluoroscopic guidance,^b a 5-Fr Pigtail angiographic catheter^c was advanced to the distal end of the abdominal aorta. Angiography was performed with 30 mL of an IV contrast agent^d at an injection rate of 15 mL per second in both dorsoventral view and left anterior 30° oblique view to demonstrate both sides of the iliac and femoral arteries, as well as the mass area. Angiographic images revealed a large, hypervascular mass of the soft tissue of the left thigh. In the early arterial phase, the angiographic images demonstrated that the mass was mainly supplied by the left deep femoral artery and left caudal gluteal artery, which appeared to be enlarged in diameter compared with equivalent arteries on the opposite side of the mass [Figure 2]. The left caudal gluteal artery was curved and displaced medially; a branch of this artery was found to be tapering and appeared obstructed at the portion distal to the hip muscles (i.e., middle and deep gluteal), suggesting tumor encroachment. The straightening and displacement of other small arterial branches surrounding the mass were also observed. Irregular vascularity was also noted within the mass; in the periphery there were considerable tumor vascularities, whereas in the central region less vascularity or avascular regions were noted. Early venous drainage from the hypervascular area appeared during the arterial phase. In the late arterial phase, tumor staining was identified as a long-lasting parenchymal opacification within tumor tissue. In the venous phase, deformed tortuous venous channels and venous lake were presented [Figures 3A–3D]. The angiographic appearance suggested the diagnosis of soft-tissue sarcoma of the left thigh based on supportive angiographic findings defined and used in human radiography. Based on these findings, transcatheter arterial embolization (TAE) was selected as the preferred treatment modality.

After diagnostic angiography, the Pigtail catheter was exchanged for a 5-Fr Cobra II catheter,^e which was used for catheterization of the arterial branches of the left deep femoral artery and left caudal gluteal artery entering the mass. Under fluoroscopic guidance, when the catheter reached one of the two target arteries supplying the mass, particles of gelatin sponge^f (1 mm × mm × 0.5 mm) mixed with 50 mL of an IV contrast agent^d and 50 mL of saline solution were injected very slowly to prevent reflux to other branches. Embolization was terminated when no further blood flow was visualized in the vessel. The Cobra II catheter was then removed and flushed completely with heparinized normal saline solution (4,000 IU of heparin in 1,000 mL of normal saline solution) to remove the particles of gelatin sponge remaining in the catheter. Subsequently, the Cobra II catheter was repositioned to the second artery, and the same procedure of TAE was repeated as in the first artery supplying the mass. Immediate follow-up angiography by manual injection of contrast agent was performed to confirm occlusion of the arteries supplying blood to the soft-tissue mass; no tumor vascularities were presented. All normal arteries surrounding the mass remained patent [Figure 4]. The catheter and sheath were removed, and the right femoral artery was depressed for 10 minutes to achieve hemostasis. The animal was then allowed to recover from anesthesia. For analgesic purposes, the dog was given carprofen^g (2 mg/kg body weight per os [PO], q 12 hours for 7 days).

On physical examination 2 weeks after embolization, the mass had decreased to 8 to 10 cm in diameter. However, the general physical condition of the dog had not improved. The dog continued to experience asthenia, anorexia, and pain. The owners declined further treatment (e.g., limb amputation) and requested euthanasia. The dog was euthanized 3 weeks after therapeutic embolization, and a necropsy was performed.

Gross pathological examination of the left thigh revealed a large, ill-demarcated fleshy tumor between the biceps femoris and semitendinosus muscles. The diameter of the mass was approximately 12 cm × 15 cm. A large amount of muscular necrosis and a hematoma were identified within the tumor. At macroscopic examination, no significant lesions were noted in the lungs. Tissue specimens from the mass were stained with hematoxylin and eosin (H & E). Histopathology demonstrated a hypercellular neoplasm that consisted of spindle cells with ovoid nuclei. Up to 12 mitoses per high-power field in 10 views were counted in the most cellular areas of the tumor. Infiltration of inflammatory cells was noted. Extensive degeneration and necrosis of musculature were also identified. Necrosis of tissue accounted for approximately 80% of the entire tumor. Histopathological examination also demonstrated neoplastic fibroblastic cells in the specimens of lung tissue. Pathological diagnosis was primary fibrosarcoma of the left thigh with pulmonary metastases.

Discussion

Fibrosarcoma accounts for approximately 35% to 45% of all soft-tissue sarcomas in the dog, based on the published literature.¹² The annual incidence of soft-tissue sarcomas in dogs is about 35 in 100,000.³ In humans, angiography is a commonly used technique for diagnosis of malignant neoplasms, and TAE is also recommended as a preoperative or palliative treatment for soft-tissue sarcomas. To the authors’ knowledge, the angiographic diagnosis of soft-tissue sarcomas and TAE have not been reported in veterinary clinical practice.

Fibrous neoplasms in the human have been extensively studied with pathology and angiography. Early in 1912, Dibbelt⁴ reported on the correlation between vascularity of fibromas and fibrosarcomas and their degree of malignancy; that is, the histopathological differentiation of the neoplastic tissue and vessels were comparable. Undifferentiated fibromatous tumors contain undifferentiated embryonic vasculature, whereas highly differentiated fibromatous tissue has normally developed vessels with differentiated wall structures. Similar conclusions about this correlation have been subsequently reached by other authors.^5–7 It is now widely accepted that there is a close relationship between the degrees of malignancy and vascularity of fibrosarcoma.

Studies in the literature confirmed the usefulness of angiography not only in demonstrating the vascular anatomy but also in revealing various features of undifferentiated embryonic vasculature of fibrosarcoma. The previous literature includes reports on angiographic findings in 79 human fibrous tumors arising from soft tissue, bone, or both; seven (9%) cases were benign and 72 (91%) were malignant.^8–14 In all seven cases of benign, fibromatous tumors (i.e., fibroma and desmoid fibroma), the only significant angiographic finding was displacement of major arteries and veins due to mass effect. The arterial supply arising from the surrounding soft tissue entering the tumor revealed a regular course pattern with tapering of the vessels; no pathological vessels or hypervascularity was demonstrated. This is in contrast to the 72 cases of fibrosarcomas, where there was considerable variability in their angiographic appearance, regardless of osseous or soft-tissue origin. Apart from displacement of vessels by mass effect, the angiographic findings in these cases revealed various characteristics related to the pathology of malignancy. The arterial blood supply to the fibrosarcomas was always derived from numerous arteries within the surrounding soft tissue. The supplying arteries usually appeared to be enlarged in diameter, likely due to high-flow hemodynamics and hyper-metabolism of malignancy. The second- and third-order arterial branches frequently revealed encasement, encroachment, or even complete obstruction because of direct invasion of malignant tumors. The circulation to the part distal to the site of obstruction was supplied through collateral arterial branches in the surrounding soft tissue as well as vascularities within the tumor. Neovascularity and hyper-vascularity were present in all of the 72 cases of fibrosarcomas reported,^8–14 of which 31 (43%) were very hypervascular, 33 (46%) were moderately hypervascular, and eight (11%) were poorly supplied with vessels. The pathological vessels usually appeared as small or large, deformed, tortuous channels, and there were alternate areas of narrowing and dilatation along the course of the vessels. Tumor stain, a long-lasting parenchymal opacification within tumor tissue, was another angiographic feature in the group of fibrosarcomas. Tumor stain mainly occurred in tumors with moderate or low degrees of vascularity, and this was theorized to be due to the lack of arterial elasticity within the pathological vessel walls, resulting in the contrast medium being held much longer in the tumor tissue than in the normal soft tissue. Angiographic studies of human fibrosarcomas demonstrated regions that contained a myriad of wide vessels in an irregular reticular pattern, while other areas showed narrowed vessels and some were avascular. Histopathological studies confirmed that the highly vascular areas corresponded to the more malignant part of the tumor; in lower grade malignant areas there was less vascularity. The avascular regions usually corresponded to areas of tumor necrosis or hematoma formation within the tumor.¹⁴

Angiographic images of the venous system in the group of human fibrosarcomas also revealed various pathological changes. The draining veins were commonly dilated and demonstrated displacement, direct invasion of the wall, complete obstruction, or intraluminal extension of tumor. In Yaghmai’s series,¹⁴ early venous drainage occurred in 66% of all 35 cases of human fibrosarcomas. Early venous drainage was usually found in those tumors with hypervascularity and thought to be due to rapid circulation. Arterio-venous shunting was also present in 12% of Yaghmai’s series.

The angiographic study in this case report revealed similar findings to those in humans, including enlargement of the arteries supplying the tumor, with tapering and evidence of obstruction suggesting tumor encroachment; irregular hypervascularity; tumor stain; tortuous venous channels in the early arterial phase; and the venous lake. These angiographic findings were supportive for the diagnosis of soft-tissue sarcoma. In the case of this report, retrieval of FNA biopsy samples and subsequent cytopathology were nondiagnostic. A likely explanation for this outcome was that as the specimens were aspirated from the central region of the tumor in this dog, they were sampling avascular areas representing regions of tumor necrosis and therefore failed to retrieve tissue allowing for a definitive diagnosis. Angiography can reveal vascularity patterns that have a close relationship with the underlying malignancy; more vascular areas represent rapidly growing and less differentiated tumor regions. Therefore, the most vascular areas identified with angiography should be the best sites for biopsy. The authors agree with Yaghmai¹⁴ that angiography done before biopsy can guide site selection for subsequent biopsy specimen retrieval, especially in cases where the diagnosis of soft-tissue sarcoma is in question. In retrospect, this would have been a valuable, preTAE procedure to have performed on the dog in this case to confirm the diagnosis of soft-tissue sarcoma, suspected based on angiographic findings.

Various treatment modalities for soft-tissue sarcomas have been reported in the dog and include surgery, radiotherapy alone or in combination with surgery, and chemotherapy. In dogs, soft-tissue sarcomas are aggressive and locally invasive and have metastatic potential; fibrosarcomas have been proven to have poor response to chemotherapy.¹⁵ Because of the tendency to develop hypoxic regions within the tumor parenchyma, soft-tissue sarcomas are considered one of the most radioresistant tumors.³ Although it was reported that soft-tissue sarcomas could be locally controlled for approximately 1 year by radiotherapy alone, with control rates of 67% at 50 Gy, the control rates decreased to only 33% after 1 year.¹⁶ Surgery has been the mainstay of treatment for soft-tissue sarcomas in dogs.¹³¹⁷ However, radical surgery (including amputation) is often necessary to achieve tumor removal and is often declined by the owner. Marginal surgery or capsulectomy, combined with postoperative radiotherapy, is often more acceptable to the pet owner. However, in soft-tissue sarcoma, there is no true capsule; the pseudocapsule consists of a layer of compressed malignant cells and normal tissue that are produced as the tumor expands. The result is that these tumors have poorly defined histopathological margins and these are often missed during the capsulectomy procedure, resulting in a high tumor recurrence rate. Radiotherapy has been used following marginal surgery to control residual disease. This should be particularly valuable in tumors of the extremities, because it permits limb sparing.¹⁸¹⁹

In humans, TAE has been widely utilized as a treatment for hypervascular tumors, such as hepatocellular carcinoma and renal cell carcinoma.^20–22 Similar to these tumors in dogs, soft-tissue sarcomas cannot be well controlled by conventional surgery, radiotherapy, or both, in human medicine. Increasing interest has therefore focused on TAE as a preoperative, adjunctive, or palliative therapy. In a report of an experimental rabbit model of VX2 sarcoma,²³ the tumors of the TAE group showed necrosis of 62%±22% of the entire tumor, compared with spontaneous necrosis of 19%±7% in the control rabbits. In one TAE rabbit, no active tumor cell could be detected on histopathology of the residual tumor. The effectiveness of TAE in the treatment of soft-tissue sarcoma was also demonstrated in a recent clinical study, in which 56% of primary and metastatic tumors showed a decrease in size after 3 months.²⁴ Generally, TAE is used as an adjunctive or palliative therapy in combination with surgery, radiotherapy, and systemic or intra-arterial chemotherapy for soft-tissue sarcomas in humans. Transcatheter arterial embolization is frequently indicated as a preoperative embolization that can significantly decrease blood loss during surgery for soft-tissue tumors. Another indication is relief of pain, since successful embolization can reduce the size of tumors, particularly in those having significant hypervascularity. However, muscle or skin necrosis, or both, might occur as a complication of TAE if the embolic agent used is small enough in size so as to embolize the terminal vessels, especially in older patients with extensive arteriosclerotic changes. Thus, the gelfoam particles larger than 200 μ are generally preferred in the procedure of TAE. It is generally accepted that establishment of collateral supply after TAE is essential to avoid the untoward necrosis. In Nagata, et al’s series of 41 patients,²⁴ one case of gluteal muscle necrosis was encountered after TAE using gelatin particles to embolize a left iliac metastatic bone tumor. In that case, arteriosclerotic obstruction of the contralateral internal iliac artery and stenosis of the ipsilateral external iliac artery were found retrospectively. Therefore, extensive arteriosclerotic changes could be a major contraindication of TAE as a treatment for soft-tissue tumors. In the dog, unlike in humans, the incidence of naturally occurring atherosclerosis is very low, and atherosclerosis has never been described as a primary vascular disease.²⁵²⁶ The other potential complications of TAE of soft-tissue tumors include fever and transient local pain and numbness in the embolized area, which usually persist for a few days after TAE. In the dog of this study, the same embolic agent (gelatin particles) was used to embolize the tumor without untoward necrosis. The tumor decreased in size 2 weeks after the procedure, and massive necrosis within the mass was confirmed histopathologically. However, the general conditions of the dog did not improve within 3 weeks following TAE. This was considered to be due to the dog’s poor condition before TAE therapy (i.e., advanced soft-tissue sarcoma with pulmonary metastases) and possible tumor lysis syndrome secondary to TAE. The fibrosarcoma in this dog was large and rich in tumor vascularities, especially in the peripheral portion of the mass, an area generally considered sensitive to TAE. With massive tumor necrosis occurring subsequently to the TAE, destruction of tumor cells might result in the continuous release of cellular breakdown products. This could compound the dog’s poor condition with asthenia and anorexia. Similar situations may also occur in humans, particularly in the patients with large hypervascular malignant tumors (e.g., primary hepatocellular carcinoma). In humans, the symptoms occasionally last >1 month. Unfortunately, the dog in this case report was euthanized on the request of the owners, and further long-term follow-up to assess the effectiveness of TAE for soft-tissue sarcoma in this dog was impossible.

Conclusion

As in humans, there appears to be a close relationship between pathological characteristics and angiographic appearances of fibrosarcoma in the dog. Angiography, therefore, could play an important role in the differential diagnosis as well as in the guidance of biopsy specimen collection. Transcatheter arterial embolization could be an effective adjunctive modality of limb sparing in veterinary medicine for fibrosarcoma or soft-tissue sarcomas. Further studies are required to better define the use of angiography as a diagnostic technique and TAE as a therapeutic modality for canine soft-tissue sarcomas.