Clinique vétérinaire

du Dr Bardet

Free proximal cortical ulnar autograft for the treatment of distal radial osteosarcoma in a dog

This report describes the use of a non-vascularized proximal cortical autograft from the ipsilateral ulna in limb-sparing surgery for the treatment of distal radial osteosarcoma. A pancarpal arthrodesis was performed to stabilize the site. Construct failure and probable local tumor recurrence were observed. The total survival time was 282 days.

Osteosarcoma (OSA) is the most common primary neoplasm of the canine skeletal system, accounting for up to 85% of bone tumors. The distal radius and proximal humerus are the 2 most common locations (1). Current treatment proto-cols are based on a combination of surgery (limb amputation or limb salvage) and adjuvant chemotherapy. There is no significant difference in survival rates between limb-salvage procedures and limb-amputation surgery when combined with adjuvant chemotherapy (1). Limb-sparing surgery is a viable alternative to limb amputation in dogs, especially if there is severe pre-existing orthopedic or neurological disease or if owners are opposed to limb amputation (1).

The purpose of this report is to describe a new limb-sparing technique involving the use of a free proximal cortical autograft from the ipsilateral ulna for the treatment of distal radial OSA in a dog. Cortical allografts (CA) have traditionally been used in limb-sparing surgeries of distal radial OSA (1). Several other limb-sparing methods have been developed in recent years. Reported techniques include pasteurized (2) or irradiated auto-grafts (3), endoprosthesis (4,5), vascularized ulnar transposition graft (6), stereotactic radiosurgery (7), and longitudinal (8) or transverse (9) bone transport osteogenesis. All these techniques are often associated with a high complication rate including infection, construct failure, and tumor recurrence (1,4).

Case description

A 4-year-old male great Dane dog (75 kg) was referred to our practice with a 2-week history of right forelimb lameness with an acute onset and an unknown origin. Lameness was obvious and the dog was weight bearing at walk, but non-weight-bearing at trot. Physical and orthopedic examination revealed an antebrachiocarpal swelling on the right forelimb, with pain on flexion and extension of the carpus. Radiography of the right forelimb revealed a mixed pattern of osteolysis and osteoproliferation on the distal third of the radius, with no apparent lesions in the distal ulna (Figure 1). The differential diagnosis for this mixed pattern bone lesion includes osteomyelitis (bacterial or fungal), metastatic bone disease, bone cysts, and primary bone tumor (1). Based on the history, signalment, lesion location, and radiographic findings, a primary bone tumor was suspected. Preoperative bone biopsy was not performed to i) avoid tumor seeding into adjacent soft tissue, and ii) decrease the risk of pathological fracture (1). Fine-needle aspiration (FNA) of the distal radial lesion was not attempted. No abnormalities were detected on preoperative 3-view thoracic radiographs, abdominal ultrasound, echocardiography, urinalysis, and blood tests [complete blood (cell) count (CBC) and biochemistry]. Treatment options discussed with the owners included chemotherapy alone and amputation or limb-sparing surgery in conjunction with chemotherapy; they elected to proceed with the latter.

Figure 1
A - Preoperative mediolateral radiograph.
B - Preoperative craniocaudal radiograph. Note the osteolysis and osteoproliferation on the distal third of the radius without apparent involvement of the ulna.

The dog was premedicated with acepromazine (Vetranquil; CEVA, Libourne, France), 0.05 mg/kg body weight (BW), IM, morphine (Cooper; Melun, France), 0.3 mg/kg BW, IM, and meloxicam (Metacam; Boehringer Ingelheim, Reims, France) 0.2 mg/kg BW, SQ. Anesthesia was induced with propofol (Diprivan; AstraZeneca, Caponago, Italy), 4 mg/kg BW, IV, to effect, and maintained under isoflurane (Isoflurane Belamont; Nicholas Piramal, Mumbai, India) and oxygen. Cephalexin (Rilexine; Virbac, Carros, France), 22 mg/kg BW, IV, was administered 30 min before induction and then every 90 min during surgery.

The dog was positioned in dorsal recumbency and a dorsal approach to the radius and carpus was performed. Caudally, the tumor was closely attached to the distal ulna. The tumor was not dissected caudally to avoid penetration. The extensor carpi radialis muscle and the common and lateral digital extensor tendons were transected proximal and distal to the tumor. An oscillating saw was used in both radial and ulnar osteotomies. The level of the transverse osteotomy of the radius, 3 cm proximal to the tumor, was determined on radiographs and confirmed by gross intraoperative evaluation (10). The ulna was osteotomized at the same level as the radius. The radius was disarticulated at the antebrachiocarpal joint and the tumor removed en bloc with the distal ulna.

The proximal aspect of the radial carpal bone was removed using an oscillating saw to provide a flat surface for pancarpal arthrodesis. The resected radial bone segment was measured and a second ulnar osteotomy performed proximally using a separate blade to yield the same length of bone, via the same dorsal approach. The length of the resected radio-ulnar segment was 9 cm, including the 3-cm free margins (this represented 35% of the total radial length). The free proximal cortical ulnar auto-graft (FPCUA) was then positioned in the radial defect. There was a slight incongruence between the FPCUA and the distal extremity of the proximal radius (Figure 2A). The FPCUA was stabilized using a 16-hole 4.5-mm dynamic compression plate (DCP; Synthes GmbH, Zuchwil; Switzerland). The plate was positioned cranially and secured to the proximal radius with 4 cortical bone screws. Four cortical bone screws were also used in the FPCUA. Five cortical bone screws were inserted distally: 1 in the radial carpal bone and 4 in the third metacarpal bone. Fifty percent of the length of the third metacarpus was covered by the plate (11). A second 15-hole 4.5-mm DCP plate (Synthes GmbH) was positioned craniolaterally to increase the rigidity of the entire construct and thus improve the stability of the FPCUA. Four cortical bone screws were used proximally in the radius and 4 cortical bone screws were used distally, 3 of which were placed in the fourth metacarpal bone, covering 30% of its length (Figure 2). Subcutaneous tissue and skin were closed routinely. The limb was protected with a modified Robert Jones bandage for 15 d. All plates were pre-bent to provide a 10° extension of the antebrachiocarpal joint.

Figure 2
A - Immediate postoperative mediolateral radiograph. Note the FPCUA stabilized with 4 screws and the slight incongruence between the proximal FPCUA and distal extent of proximal radius (arrow).
B - Immediate postoperative craniocaudal radiograph.

Postoperative analgesia consisted of subcutaneous morphine (0.5 mg/kg, BW, SQ every 4 to 6 h, as needed) and oral meloxicam (Metacam 2.5; Boehringer Ingelheim, Reims, France), 0.1 mg/kg BW, q24h for 48 h. The dog was discharged, with oral tramadol for 7 d (Tramadol LP Sandoz), 4 mg/kg BW, PO, q8 to 12 h; and oral meloxicam (same dosage) for 15 d. Oral cephalexin (Rilexine 600 mg; Virbac), 22 mg/kg BW, q8h was administered for 2 wk and exercise was restricted to short walks on a leash for 8 wk. The dog was re-examined 2, 3, 6, and 9 mo after the surgery. Thoracic and limb radiographs and limb function were evaluated at each examination.

The histopathological diagnosis of the excised bone con-firmed a grade II fibroblastic osteosarcoma (OSA) without ulnar involvement, but soft tissue extracompartmental presence (stage IIB). Surgical resection was considered complete with no evidence of neoplasia at the radial osteotomy site. An alternating protocol of doxorubicin (30 mg/m2) and carboplatin (300 mg/m2) was planned every 3 wk for a total of 6 treatments (12). The first treatment was started 2 wk after surgery. A CBC was taken 2 wk after each carboplatin administration; serum biochemistry was performed every 2 mo, and an echocardiogram was done during the visit of the last treatment. There was no evidence of gastrointestinal upset, myelosupression, or cardio-toxicity during or after chemotherapy.

Lameness was subjectively graded as absent, mild, moderate, or severe depending on the degree of weight-bearing on the operated limb. Lameness was graded as mild if weight-bearing and not present at walk or trot; moderate if the lameness was weight-bearing and present at all paces; and severe if non-weight bearing. Limb function was considered excellent if lameness was absent, good with mild lameness, and poor if lameness was either moderate or severe (3,13). Lameness progressively improved from toe-touching 2 d after surgery to full weight-bearing without visible lameness at walk or trot 2 mo after surgery. Three months after surgery lameness was still mild with good limb function, without any significant changes on limb palpation and on limb and thoracic radiographs. Six months after surgery, although lameness was still mild with good limb function, some signs of early construct failure were observed. The most distal and the most proximal screw on the third metacarpal bone were loose and a radiolucent line was observed around all 4 screws in the FPCUA, indicating micromotion. Moderate osteoproliferation was observed on the plate on the proximal radius and some cortical resorption was observed on the distal third of the FPCUA, suggesting an osteomyelitis (Figure 3). A FNA with a culture was not done because there were no detectable signs of infection in the operated limb (no fistulas or soft tissue swelling). Implant revision was not per-formed due to good limb function.

Figure 3
A - 6-month postoperative (PO) mediolateral radiograph. There is loosening of the most distal and the most proximal screws on the third metacarpal bone. Note the radiolucent line around the 4 screws of the FPCUA (arrowheads).
B - 6-month PO craniocaudal radiograph. Note the mild osteoproliferation on the proximal radius (arrows).

Nine months after surgery, the dog was admitted with a 2-day history of progressive weakness. Abdominal ultrasound revealed effusion, which was confirmed as a hemoabdomen on abdominocentesis. The spleen was enlarged with several hypoechoic nodules scattered throughout the splenic and hepatic parenchyma. On exploratory laparotomy, a 4-cm splenic nodule was found to have ruptured. Multiple nodules were disseminated in all hepatic lobes. Splenic haemangiosarcoma (HSA) with hepatic metastasis was diagnosed on histological examination following splenectomy and hepatic biopsy. Thoracic radiography and palpation of regional lymph nodes did not provide evidence of metastases. Echocardiography was within normal limits. Doxorubicin chemotherapy (30 mg/m2, IV) was begun 1 wk after surgery. A whole fresh blood transfusion was performed, and the dog was discharged 3 d after the splenectomy.

Prior to this episode, the owner told us that lameness was not present at walk, and they considered that the dog had good limb function. Osteoproliferation radiating from the proximal radial cortex had been progressing in a sunburst pattern, suggestive of local tumor recurrence or a severe osteomyelitis (Figure 4). The FPCUA was more radiolucent and 2 screws had become looser (the second and third screw in the FPCUA from proximal to distal; Figure 4). Implant revision or diagnostic surgery was not considered on account of the poor prognosis associated with the splenic and hepatic HSA and in light of the good limb function. Two weeks after the splenectomy, the dog was readmitted for recurrent hemoabdomen, presumably following the rupture of a hepatic nodule. The dog was euthanized. The owners declined necropsy. The total survival time was 282 d.

Figure 4
A - 9-month postoperative (PO) mediolateral radiograph. Note the loosening of the second and third proximal FPCUA screws (arrowheads). B - 9-month PO craniocaudal radiograph. There is marked osteoproliferation (sunburst pattern; arrows), compatible with tumor recurrence or osteomyelitis.


The use of a free cortical ulnar autograft from the mid-ulnar diaphysis has been described previously to stabilize a segmental mandibulectomy performed to remove an ossifying epulis (14). The distal aspect of the ulna has also been used successfully as a free (15) or rotational vascularized cortical autograft (6). To the authors' knowledge, this is the first description of a non-vascularized proximal cortical autograft from the ipsilateral ulna for limb-sparing surgery of a distal radial OSA.

The advantages of FPCUA include the ready availability of the graft with no need for special equipment, bone-banking facilities, or external beam radiation therapy. Dogs with pre- or intra-operative evidence of tumor extension into the distal ulna can still be good candidates, since the technique allows the en bloc removal of the distal ulna. Liptak et al (4) demonstrated that preservation of the ulna is not required in an endoprosthesis model for limb-sparing of the distal radius for improved stability in axial compression. Whereas some techniques require cumber-some postoperative management (longitudinal and transverse bone transposition), FPCUA was easier to manage in this case. Furthermore, the length of the FPCUA can be adjusted to fit the radial segment, providing there is sufficient remaining ulnar length after ulnar ostectomy. The lack of vascularization could have made the graft more susceptible to infection, but we did not observe any physical sign of infection on the FPCUA, despite 6- and 9-month radiographic signs. The 9-month post-operative radiographs showed intense periosteal remodelling with a sunburst periosteal pattern, which is more typical of a neoplastic etiology. However, we cannot exclude osteomyelitis. A FNA or a bone biopsy and a culture should have been done, but they were not performed at that time because of the good limb function and because there was no external sign of infection.

Accurate preoperative assessment of the proximal extent of the tumor is essential when selecting candidates for limb-sparing; the best candidates are those with small tumors (, 40% of radius involvement including 3-cm tumor-free margins, as it was in our case). Radiography can be an accurate method for measuring the margins (10); however, a computed tomography (CT) scan could have been useful for determining tumor length. An FPCUA requires permanent internal fixation, because the non-vascularized autograft will never be fully incorporated. Allograft non-union is relatively common in canine limb-sparing surgery. The clinical impact of non-union is usually minimal since the allograft-host bone interface is rigidly stabilized and protected with a plate (1).

Construct failure in this case was considered to be a minor complication as surgical revision was not necessary. It was first observed 6 mo after surgery in 6 screws, without any sign of lameness. Three months later, 2 screws had become looser, but the dog remained asymptomatic. Revision was not considered necessary because of good limb function. A second plate was added craniolaterally to decrease stress concentration on the FPCUA. We covered 50% of the third metacarpal bone in this pancarpal arthrodesis (11); up to 80% of the metacarpal bone should be covered to minimize the risk of instability (6). En-bloc resection of the ulna and radius did not have a negative impact on construct stability in 1 study (4). The use of polymethyl methacrylate (PMMA) could have increased the strength of our autograft. In a study using intercalary bone grafts, implant failure was significantly associated with non-cemented allografts (13). The use of locking compression plates could also have improved construct stability.

A preoperative biopsy was not taken because the history, age, and breed, and radiographic findings were characteristic of a primary bone tumor and not typical of either metastatic neoplasia or fungal or bacterial osteomyelitis. Furthermore, the owner's decision regarding surgery would not have been influenced by the knowledge of tumor type. Intraoperative or preoperative cytology could have been performed to confirm diagnosis before proceeding with limb-sparing surgery. Distal radial OSA rarely affects the distal ulna (1); however, the neoplastic pseudocapsule frequently adheres to the latter. We made the decision to use the proximal ulnar bone preoperatively, because the length of the radial tumor was such that this technique gave 3-cm margins. If the tumor had been longer this technique would not have been feasible with free radial tumor margins. In our case, the tumor length with free margins was 35% of the total radial length. A resected radial length over 40% could jeopardize the FPCUA, because the proximal ulnar osteotomy could be too close to the elbow stabilizers such as the collateral ligaments and the insertion tendons of the biceps brachii and brachialis muscles, to counteract the strong pull of the triceps muscle, increasing the probability of elbow luxation. Tumor control was good with a disease-free interval of 282 d, although local tumor recurrence was strongly suspected. Radiographs taken 9 mo after surgery showed a sunburst osteoproliferation on the proximal radius suggestive of local tumor recurrence. The owners did not give their consent for a necropsy and we did not attempt an antemortem FNA, so we were unable to confirm this suspicion.

The limitations of this case report are that limb function assessment was subjective and there was a lack of diagnosis of the osteoproliferative lesion. Limb function is difficult to assess clinically, as we can expect some alteration of normal gait resulting from the pancarpal arthrodesis.

In conclusion, a free proximal cortical ulnar autograft could be effective in limb-sparing surgery in distal radial OSA, pro-viding the tumor is small enough to allow the proximal ulnar autograft. Further evaluation of this technique is required before it can be recommended routinely for limb-sparing in dogs.


1. Dernell WS, Ehrhart N, Straw RC. Tumors of the skeletal system. In: Withrow SJ, MacEwen EG, eds. Small Animal Clinical Oncology, Philadelphia, Pennsylvania: Saunders, 2007:541–570.

2. Buracco P, Morello E, Martano M, et al. Pasteurized tumoral autograft as a novel procedure for limb-sparing in the dog: A clinical report. Vet Surg 2002;31:525–532.

3. Liptak JM, Dernell WS, Lascelles BDX, et al. Intraoperative extracorpo-real irradiation for limb sparing in 13 dogs. Vet Surg 2004;33:446–456.

4. Liptak JM, Ehrhart N, Santoni BG, et al. Cortical bone graft and endoprosthesis in the distal radius of dogs: A biomechanical comparison of two different limb-sparing techniques. Vet Surg 2006;35:150–160.

5. Liptak JM, Dernell WS, Ehrhart N, et al. Cortical allograft and endo-prosthesis for limb–sparing surgery in dogs with distal radial osteosarcoma: A prospective clinical comparison of two different limb-sparing techniques. Vet Surg 2006;35:518–533.

6. Pooya HA, Séguin B, Mason DR, et al. Biomechanical comparison of cortical radial graft versus ulnar transposition graft limb-sparing techniques for the distal radial site in dogs. Vet Surg 2004;33:301–308.

7. Farese JP, Milner R, Thompson MS, et al. Stereotactic radiosurgery for treatment of osteosarcoma involving the distal portions of the limbs in dogs. J Am Vet Med Assoc 2004;10:1567–1572.

8. Ehrhart N. Longitudinal bone transport for treatment of primary bone tumors in dogs: Technique description and outcome in 9 dogs. Vet Surg 2005;34:24–34.

9. Jehn CT, Lewis DD, Farese JP, et al. Transverse ulnar bone transport osteogenesis: A new technique for limb salvage for the treatment of distal radial osteosarcoma in dogs. Vet Surg 2007;36:324–334.

10. Leibman NF, Kuntz CA, Stein PF, et al. Accuracy of radiography, nuclear scintigraphy and histopathology for determining the proximal extent of distal radius osteosarcoma in dogs. Vet Surg 2001;30:240–245.

11. Whitelock RG, Dyce J, Houlton, JE. Metacarpal fractures associated with pancarpal arthrodesis in dog. Vet Surg 1999;28:25–30.

12. Bacon NJ, Ehrhart NP, Dernell WS, et al. Use of alternating administra-tion of carboplatin and doxorubicin in dogs with microscopic metastasis after amputation for appendicular osteosarcoma: 50 cases (1999–2006). J Am Vet Med Assoc 2008;232:1504–1510.

13. Liptak JM, Dernell WS, Straw RC, et al. Intercalary bone grafts for joint an limb preservation in 17 dogs with high-grade malignant tumors of the diaphysis. Vet Surg 2004;33:457–467.

14. Bracker KE, Trout, NJ. Use of a free cortical ulnar autograft following en bloc resection of a mandibular tumor. J Am Anim Hosp Assoc 2000; 36:76–79.

15. Szentimrey D, Fowler D, Johnston G, et al. Transplantation of the canine distal ulna as a free vascularized bone graft. Vet Surg 1995;24: 215–225.