History

A 5-year-old 43.0-kg (94.6-lb) castrated male Rottweiler was evaluated for a second opinion because of a chronic draining tract along the medial aspect of the right tibia. Bilateral cranial cruciate ligament ruptures had been diagnosed approximately 14 months earlier, and the dog had undergone a left tibial plateau leveling osteotomy (TPLO) followed by a right TPLO approximately 14 and 10 months earlier, respectively. In addition, 5 months before being evaluated at our facility, the dog was examined by the primary care veterinarian for signs including a semi-firm swelling and a draining tract along the medial aspect of the right stifle joint that were presumed to be the result of a draining abscess, consistent with a surgical site infection. The dog was empirically treated with marbofloxacin (4.1 mg/kg [1.9 mg/lb], PO, q 24 h for 10 days) and pain medications, including gabapentin (6.2 mg/kg [2.8 mg/kg], PO, q 24 h) and carprofen (no dosage provided). Because there was no improvement after 8 weeks of antimicrobial treatment, bacterial culture and susceptibility testing were performed on a sample of the draining fluid. Results indicated methicillin-resistant Staphylococcus pseudointermedius that was susceptible to doxycycline, chloramphenicol, and amikacin. The dog underwent a right TPLO explant procedure following the culture results. No additional culture was performed at the time of surgery. After surgery, administration of amikacin (15.0 mg/kg [6.8 mg/lb], SQ, q 24 h) was initiated. During the following 3 months and before referral to our hospital, the dog was examined by the primary care veterinarian 5 times for continued drainage along the medial aspect of the right tibia. A second bacterial culture and susceptibility testing were performed during that time, and results were unchanged. Therefore, treatment with amikacin was continued.

On initial examination at our hospital, the dog had mild weight-bearing lameness (grade 1/5 to 2/5¹) in its left hind limb and would off-load the left hind limb at a stance. The left hind limb had no evidence of a draining tract; however, 2 tracts draining seropurulent fluid were noted along the medial aspect of the right tibia. The more proximal draining tract was located at the distal aspect of the previous surgical incision, and the other draining tract was located approximately 2 to 3 cm more caudally and distally. No signs of pain were elicited with movement of the right stifle joint through its range of motion or with long bone palpation of the right hind limb. Results of a CBC and serum biochemical analyses were unremarkable, and orthogonal radiographic views of the right stifle joint were obtained (Figure 1).

View Full Size

Mediolateral (A) and craniocaudal (B) radiographic views of the right stifle joint of a 5-year-old 43.0-kg (94.6-lb) castrated male Rottweiler that was evaluated for a 5-month history of 2 draining tracts following tibial plateau leveling osteotomy and subsequent implant removal and antimicrobial treatment.

Citation: Journal of the American Veterinary Medical Association 254, 9; 10.2460/javma.254.9.1045

• Download Figure
• Download figure as PowerPoint slide

Determine whether additional imaging studies are required, or make your diagnosis from Figure 1—then turn the page →

Radiographic Findings and Interpretation

Radiography of the right stifle joint revealed well-demarcated, smoothly margined lysis along the caudal and medial aspect of the tibial diaphysis and proximal metaphysis (Figure 2). Increased opacity of the tibia along the area outlining the region of bony lysis was evident, and there was a narrow zone of transition between the margin of bony lysis and the appearance of normal cortex. On the craniocaudal radiographic view, an ill-defined soft tissue opacity was evident adjacent to the area of bony lysis. On the mediolateral radiographic view, caudal displacement of the popliteus and gastrocnemius muscles created a mass effect caudal to the proximal aspect of the tibia. In both views, an irregular margin of the tibia, suggestive of a periosteal reaction, was evident just proximal and distal to this mass effect.

View Full Size

Same radiographic images as Figure 1. There is evidence of bony fusion from the previous tibial plateau leveling osteotomy with mild thickening of the patellar ligament at its insertion (long arrow; A). Also evident are osteolysis with subchondral increased opacity (short arrows; A and B) along the caudal and proximal aspects of the tibia and a mass effect (outlined by asterisks; A and B) in the adjacent soft tissue that displaces the popliteus and gastrocnemius muscles caudally. Additionally, irregular margins of the tibia at the proximal and distal aspect of this mass effect are suggestive of a periosteal reaction.

Citation: Journal of the American Veterinary Medical Association 254, 9; 10.2460/javma.254.9.1045

• Download Figure
• Download figure as PowerPoint slide

The combined presence of cortical bony lysis and an irregular periosteal reaction suggested an aggressive bony lesion. The differential diagnosis list for these findings and the mass effect included a soft tissue neoplasia with secondary soft tissue or bony infection and an abscess with secondary osteomyelitis from a chronic implant-associated infection and possible antimicrobial resistance. Given the dog's history of long-term infection and antimicrobial treatment, the differential diagnosis of abscess formation with secondary osteomyelitis was prioritized, and the potential presence of a foreign body (eg, gauze) was not considered, although placement of gauze caudal to the proximal aspect of the tibia to protect the popliteal artery during the TPLO procedure has been described.²

Treatment and Outcome

The dog was anesthetized and underwent surgical exploration of its draining tracts. The distal draining tract was incised and followed to a communication with the proximal draining tract. The underlying subcutaneous tissue was incised, as was fibrous tissue that was encountered. The incision was extended proximally for better visualization. The pes anserinus was elevated to expose the proximal aspect of the tibia. Abundant fibrous tissue was encountered caudal to the tibia in the region of the popliteus muscle. Blunt dissection revealed an intact, nonradiopaque gauze sponge located medially and caudally to the proximal aspect of the tibia. The gauze sponge was extracted, and samples of the deep subcutaneous tissue adjacent to where the gauze had been were submitted for bacterial culture and susceptibility testing.

The dog recovered uneventfully and was discharged the following morning. Results of bacterial culture and susceptibility testing indicated methicillin-resistant S pseudointermedius that was susceptible to doxycycline, chloramphenicol, and amikacin. Given the dog's history of extended treatment with amikacin, treatment with amikacin was discontinued and chloramphenicol (41.8 mg/kg [19.0 mg/kg], PO, q 8 h for 6 weeks) was initiated along with recheck CBCs every 2 weeks until chloramphenicol was discontinued.

A final evaluation was performed 7 weeks after surgery, which was also 7 days after the last dose of chloramphenicol was administered. At the recheck examination, the owners reported that the dog had no signs of right hind limb lameness or incisional discharge. Clinical evaluation confirmed the same.

Comments

Gossypiboma is the term used for a granulomatous inflammatory reaction caused by cotton, most commonly a retained surgical sponge, towel, or gauze,³ left in a patient after surgery. There are 2 types of foreign-body reactions that may result from the presence of retained surgical gauze: an aseptic granulomatous reaction that may result in delayed or no clinical signs or an exudative reaction that often results in the formation of a draining sinus and subsequent infection.³ Radiographic features of a retained surgical sponge may include a localized whirl-like gas pattern, mass effect, peripheral calcification, and bony lysis.^3,4

In the dog of the present report, differential diagnosis required consideration of a mass effect combined with bony changes. The presence of a distinct zone of transition to normal bone suggested a non-aggressive lesion. On the other hand, indications of cortical destruction and irregular periosteal reaction suggested an aggressive lesion such as neoplasia or osteomyelitis.⁵ In hindsight, gossypiboma could have explained the radiographic findings and the dog's lack of response to long-term antimicrobial treatment.

In addition to radiography, other diagnostic imaging modalities (MRI, CT, or ultrasound) could have been considered to help narrow the differential diagnosis list and identify a treatment plan. In humans, T2-weighted MRI findings for retained gauze may include a well-defined wall with a whorled internal structure.⁶ Findings on CT for gossypiboma in humans include a foreign body with a whirl-like, wavy, striped, or spotted internal structure, which is not consistent with the presence of an abcess.^7,8 Ultrasonography may reveal less specific characteristics, including the presence of a well-demarcated primarily hypoechoic mass with echogenic material.⁸ Further, fine-needle aspiration has also been used to diagnose gossypiboma in dogs.³

In human medicine, risk factors favoring retention of surgical sponges have been identified and include emergency surgery, a change in the intended surgical procedure, and patient obesity.⁹ Further research is required to determine whether similar risk factors apply in veterinary medicine.

Efforts should be made to reduce risks of surgical sponges being left in patients because retained sponges may have a clinical impact on patient morbidity and death. In a case series,³ 2 of 8 dogs died from complications of sponge retention. Delayed complications have also been reported, including 3 reports^10–12 of osteosarcoma, 2 instances of which were identified caudal to the proximal aspect of the tibia years after surgical correction of a cranial cruciate ligament tear.

Gossypiboma is a rare, preventable complication. We suggest that veterinarians use radiopaque gauze and count gauze sponges at the beginning and end of surgery. Both of these preventative measures are standard practice in human medicine,⁹ and given the subtle radiographic signs of a gossypiboma,³ use of radiopaque gauze could be particularly useful for practices in which radiographic interpretation is primarily performed by practitioners who are not board-certified veterinary radiologists. In addition, findings in the dog of the present report underscored that practitioners should consider gossypiboma as a differential diagnosis in patents with a history of surgery and appropriate suggestive findings on physical and radiographic examinations.