Osteosarcoma at the Site of a Triple Pelvic Osteotomy in a Dog

Introduction

Of all skeletal neoplasms, osteosarcoma (OSA) is the most common primary bone tumor in dogs, representing 3% to 6% of all canine neoplasms.¹ Osteogenic sarcomas can be subdivided into those that occur spontaneously and those that occur in association with a previous bone fracture. Spontaneously occurring OSAs often involve the metaphyseal region of the appendicular skeleton.^2–4 Fracture-associated OSAs also affect the appendicular skeleton and differ from spontaneous forms in their location and frequency.^3,4 Fracture-associated OSAs occur more frequently in the diaphysis and represent only a fraction of all OSA cases reported.^3,4 Although reports of fracture-associated sarcomas have been limited to traumatic fractures, two recent case reports of canine OSA following total hip arthroplasty suggested that these tumors may have formed secondary to chronic loosening of orthopedic implants.^5,6 The history of chronic loosening of orthopedic implants supports the hypothesis that altered cellular activity, along with recurrent regeneration and repair from an unstable defect in bone, may lead to the formation of an OSA. This report describes an OSA in the ilial body of a dog diagnosed 11 years after a triple pelvic osteotomy (TPO) was performed.

Case Report

A 12-year-old, 34.5-kg, spayed female golden retriever was presented to the University of Minnesota Veterinary Medical Center with a 12-week history of nonweight-bearing lameness of the right hind limb. At 9 months of age, the dog underwent an elective ovariohysterectomy and radiographic evaluation for canine hip dysplasia. On orthopedic examination, bilateral positive Ortolani signs were palpated; however, no pain was elicited on flexion or extension of either coxofemoral joint. Neurological examination was within normal limits. Radiographic evaluation of the coxofemoral joints revealed no degenerative joint disease, despite a high degree of subluxation in both joints [Figure 1]. Frog-leg views of the pelvis confirmed good seating of the femoral heads in the acetabula. A TPO was recommended for the right side because of the more severe subluxation on the right side. A right TPO was performed and stabilized using a 45°, seven-hole, 2.7-mm Synthes^a plate for the ilium and a 22-gauge hemicerclage wire for the ischium. Postoperative radiographs showed a well-positioned plate with screws extending into both cortices of the ilial shaft and excellent acetabular coverage of the right femoral head. Recovery from surgery was unremarkable.

Six weeks postoperatively, the dog was using the operated leg well. Radiographs revealed considerable reduction in the degree of subluxation of the right coxofemoral joint. The ilial osteotomy was healing adequately with minimal callus formation, and all surgical implants appeared stable. The left coxofemoral joint also had less subluxation and minimal progression of osteoarthritis. Little muscle atrophy was noted in either hind limb. A TPO was recommended at this time to correct the left coxofemoral subluxation; however, further surgical intervention was not pursued.

Radiographs taken 12 weeks postoperatively revealed complete healing of the right ilial shaft and ischial plateau [Figure 2]. The right coxofemoral joint subluxation was well-reduced; however, there was increased progression of the degenerative joint disease in the left coxofemoral joint.

Eleven years after the right TPO was performed, the dog was presented for right hind-limb lameness of 12 weeks’ duration. Orthopedic examination revealed pain on extension and flexion of the right coxofemoral joint. No neurological deficits were noted. Radiographs demonstrated an aggressive bone lesion of the right ilium with profound periosteal proliferation and punctate lysis that extended along the ilium caudally and into the right ischium [Figure 3]. The epicenter of the bony lesion was at the level of the right TPO plate. Thoracic radiographs were not taken at this time. The owners elected to manage the dog’s pain with morphine sulfate^b (0.43 mg/kg per os [PO] q 12 hours) and carprofen^c (2.2 mg/kg PO q 12 hours), and the dog was presented 2 days later for euthanasia.

At necropsy, a fleshy, cream-colored mass (4.5 × 2.5 cm) occupied the body of the rectus femoris muscle, immediately dorsal and cranial to the right acetabulum, extending cranioventrally to the right femoral head. The bone of the right femoral head and neck region was slightly irregular. The entire right side of the pelvis, especially the ischium, was markedly thickened and firm with irregular margins [Figure 4]. Multifocal, hard, whitishtan nodules were observed in the lungs, heart (myocardium), liver, spleen, both kidneys, and the right external iliac lymph node. These nodules ranged from 0.2 to 10 cm in diameter and could not be incised with a knife.

For histology, samples of lung, liver, kidney, spleen, heart, right ilium, right external iliac lymph node, and right rectus femoris skeletal muscle were collected. Sections were decalcified, cut at 4 μm, and stained with hematoxylin and eosin (H&E). Sections of the medial surface of the ilium were composed of lobules of pleomorphic, neoplastic, polygonal to spindle-shaped cells separated by variable amounts of collagen, interconnecting seams of eosinophilic material (interpreted as osteoid), and mineral deposits [Figure 5]. The tumor cells had round, hyperchromatic to vesicular nuclei with clumped chromatin and small single or multiple nucleoli. The nuclear to cytoplasmic ratio was approximately 1:1, and the cell borders were relatively distinct. The tumor cells exhibited varying degrees of anisocytosis and anisokaryosis with occasional mitotic figures. Scattered throughout the histological sections were extensive osteoid trabeculae with relatively few tumor cells. The skeletal muscle adjacent to the mass in the right rectus femoris muscle was replaced by sheets of neoplastic mesenchymal cells (i.e., osteoblasts) with large, multifocal areas of hemorrhage and necrosis. Many deposits of osteoid were frequently noted in the areas of neoplastic infiltration. Metastases with histological features similar to the pelvic mass, including numerous osteoid trabeculae, were found in the lung, heart, liver, kidneys, spleen, and right external iliac lymph node.

Discussion

Fracture-associated sarcomas have been identified as one of the principal factors in OSA formation in the appendicular skeleton of dogs.^2–4,7,8 The frequency and location of fracture-associated sarcomas in long bones differ from those that arise spontaneously.^3,4 Spontaneous OSAs are present in the metaphyseal region of long bones 95% of the time, with the majority occurring in the distal radius.³ In contrast, fracture-associated sarcomas occur in the diaphyseal region 85% of the time, with the femur being overrepresented.^3,4,8 Although there are multiple reports of OSA associated with fracture sites and other traumatic injuries to the appendicular skeleton, relatively few reports involve the axial skeleton. In the case reported here, a variety of factors may have led to the formation of the ilial OSA, including age, size, breed of the dog, the surgical osteotomy site, and the metallic plate and screws.^2–17 The pathogenesis of the tumor in this case was unclear; however, the location of the mass directly over the TPO site, the lack of tumor involvement in other skeletal sites (such as the contralateral side of the pelvis), and the rarity of spontaneous OSAs in the pelvic region made it less likely that the tumor arose spontaneously.

Most of the bony mass in this dog’s pelvis was centered directly over the TPO plate and osteotomy site. Given the close proximity to the surgical site, several inciting factors may have been involved, such as excessive tissue damage, altered cellular activity, corrosion of the metallic implant, metallic wear particles from the implant, inflammation, or infection.^3,4,7 The history of this dog revealed that the lameness did not occur until 12 weeks prior to euthanasia, and prior radiographs indicated a well-seated plate and screws, suggesting that movement of the surgical plate and screws was not a likely factor in the development of the OSA in this case. Histopathological sections revealed no evidence of chronic inflammation or infection, which contrasted with other reports of fracture-associated sarcomas.³

Reports of OSAs in humans have revealed that abnormal cellular activity is often one of the predisposing factors in carcinogenesis.^3,18 The abnormal cellular activity manifests as chronic regeneration and repair, occasionally resulting in neoplastic growth.^3,18 In a study of 552 cases of human OSA, 13% of the bone alteration was a result of preexisting disease; 6% was from Paget’s disease; 4% occurred in irradiated bone; 2.5% arose from hereditary multiple exostosis; and 0.7% was from polyostotic fibrous dysplasia.¹⁸ Although it has been shown that cellular activity can be altered by various disease processes, surgical implants and their metallic components may also induce tumors in both humans and animals.^{5,6,9–16,19}

Implanted metallic particles and corrosion products from cobalt, cobalt-chromium-molybdenum alloys, cadmium, and nickel were examined in mice and were shown to induce malignant neoplasms.¹⁹ Similar findings were found in rats when sheared particles from metal-on-metal prostheses, such as the McKnee total hip and the Shier and Waldius chrome-cobalt hinge knee prostheses, were injected into rat muscle.^17,19 Since implantation of metallic particles and their corrosion products has been shown to cause tumors in rodents, this suggests that implantation of these same materials in the form of metallic prostheses could also result in tumor development. However, the presence of these metallic particles in the tissues neither proves nor disproves carcinogenicity.¹⁴

Fracture-associated sarcomas are thought to arise from disruption of the natural healing processes; however, these tumors generally have a prolonged latency period.³ A 20-year delay is usually seen in humans, whereas dogs seem to have a shorter mean interval of 5.5 years between fracture and tumor diagnosis.^3,4 The dog presented in this report had an 11-year interval from the date of the surgical procedure (osteotomy) to radiographic diagnosis of OSA.

Although this dog had no previous history of a pelvic fracture, it did have a prior osteotomy. Since it is recognized that fracture sites can be associated with sarcoma formation, one can hypothesize that osteotomy sites may have a similar risk for tumor development, because an osteotomy could be described as an iatrogenic fracture. Keel et al. reported an OSA in a human with a history of an osteotomy to correct a genu valgum deformity.¹² The tumor in this person was thought to be associated with the placement of staples over the osteotomy site, but the OSA could also have arisen from the osteotomy itself.¹² Numerous theories surround the development of osteogenic sarcomas, including an association with metallic implants, sites of previous fractures, the site of an osteotomy (as in the present case), or a combination of these factors.^2–17

Conclusion

The signalment and clinical signs of the dog presented in this case were similar to those of other dogs with OSA; however, in this case, the mass was located directly over a stable TPO plate placed 11 years earlier. While it was not possible to prove or disprove the association of this tumor and the prior osteotomy, the possibility should be considered. With more young dogs receiving elective orthopedic surgical procedures such as TPOs, tibial plateau leveling osteotomies, and other corrective osteotomies, this case raises questions of how OSA and other malignant tumors form under stable fixation, and it heightens the clinician’s awareness of the potential associations between osteotomies and OSA development.

Acknowledgments

The authors thank Dr. Kari Anderson for assistance with reviewing the radiographs; they are also grateful to Dr. David Hayden and Dr. Al Lipowitz for technical assistance and for reviewing the manuscript.