Abstract

A three-year-old 8.2 kg female green sea turtle (Chelonia mydas) was referred to the Veterinary Medical Teaching Hospital, University of Florida for evaluation of chronic bilateral polyarthritis of the front flippers. Five weeks prior to presentation, radiographic evaluation of the right forelimb detected an abnormality at the elbow joint. Cultures of blood and joint fluid were positive for a gram-negative organism. The turtle was treated by the referring facility with amikacin (5 mg/kg i.m. initial dose, then 2.5 mg/kg i.m. q 72 hr x 10 day), enrofloxacin (5 mg/kg i.m. q 48 hr x 10 day), cefotaxime (20 mg/kg i.m. s.i.d. x 10 day), and penicillin G procaine (10,000 IU/kg s.i.d. x 5 day, then q 48 hr), with no significant improvement in clinical signs. Upon presentation, physical examination confirmed enlargement of the joints in both front flippers with impairment of function. Hematologic parameters were within normal range and serum chemistries displayed hyperproteinemia (8.0 mg/dl) and hyperuricacidemia (12.2 mg/dl). Radiographs revealed multiple areas of bony lysis in the proximal aspect of both humeri. Lesions were most severe on the right side, with bony lysis observed within the right scapula adjacent to the shoulder joint, distal aspect of the right humerus at the level of the elbow joint, and radiolucency, irregularity, and bony lysis identified within the second carpal bone and second carpometacarpal joint. Histopathology and culture of bone biopsies obtained from both humeri identified chronic severe osteomyelitis due to Serratia marcescens. Blood cultures were negative for growth at two and five days. Parenteral therapy with amikacin (Amiglyde-V, 250 mg/ml, Fort Dodge Laboratories, Fort Dodge, Iowa, USA) (5 mg/kg i.m. initial dose, then 2.5 mg/kg i.m. q 72 hr) was initiated based upon the susceptibility pattern.

The chronic nature of the lesion and results of diagnostic testing warranted surgical intervention for removal of infected material combined with localized antibiotic therapy using antibiotic-impregnated polymethyl methacrylate beads. At surgery, a thick fibrous capsule containing caseous necrotic material was identified and removed from the left and right scapulohumeral joint and right elbow joint. Following removal of infected material, gentamicin-impregnated polymethyl methacrylate beads strung on surgical wire were implanted into the defect. Blood samples were collected from the occipital sinus every 24 hr for 16 days following surgery for evaluation of systemic gentamicin levels.

Radiographic evaluation performed three weeks postoperative demonstrated no appreciable change in the previously identified lytic lesions. Additionally, bony lysis was observed in both acetabula and femoral heads with poorly defined sclerosis of adjacent pelvic bone. Bilateral coxofemoral luxation was also present. Appetite and behavior had begun to deteriorate and the turtle showed little ability to use the front flippers effectively. Due to the appearance of additional sites of septic arthritis and poor prognosis for recovery, the turtle was euthanized. Postmortem findings confirmed Serratia marcescens as the causative agent of the septic arthritis. Additionally, a large granuloma containing gram-negative rods was identified within the cerebrum. Fluid was obtained postmortem from implanted joints for evaluation of local gentamicin levels.

In domestic species, osteomyelitis results from infection of bone through exogenous contamination or hematogenous exposure. Potentiating factors in the pathogenesis of osteomyelitis include massive contamination, tissue ischemia, foreign material, highly virulent pathogens, and suboptimal immune function. Because there was no history of previous surgical intervention or trauma in this patient, primary infection and spread was assumed to occur through hematogenous exposure. Although negative results were obtained on blood culture, bacteremia associated with osteomyelitis is intermittent and antibiotic therapy administered prior to referral may have inhibited (without eliminating) other foci of bacterial infections. Treatment of chronic osteomyelitis is difficult and prolonged and, in cases where a definitive source of infection cannot be identified, prognosis is guarded.

The rationale for application of antibiotic impregnated polymethyl methacrylate beads (AIPMMA) is to provide locally therapeutic concentrations of antibiotics. Because there is little systemic absorption of drug, the potential for toxic reactions is decreased. This technique is particularly useful against infections localized in tissues with poor drug penetration, in situations where long-term or systemic use of antibiotics is contraindicated, or with intractable or dangerous patients. AIPMMA beads are made using commercially available polymethyl methacrylate (PMMA) compounds (Surgical Simplex, Howmedica, Rutherford, NJ, or Bone Cement, Zimmer, Charlotte, NC). Antibiotics are mixed thoroughly with the copolymer powder prior to the addition of the liquid monomer. Antibiotics chosen should be bactericidal, effective against the pathogen of interest, therapeutic at low concentrations, water soluble, and demonstrate low tissue toxicity. Because the polymerization reaction is exothermic, antibiotics used must also be heat stable up to 100°C. In this clinical case, gentamicin was chosen based upon its low economic investment, effectiveness at low concentrations, water solubility, heat stability, and susceptibility of the bacterial pathogen.

Gentamicin levels in serum and joint fluid were assayed using a fluorescence polarization immunoassay (Texas A&M University, College Station, TX, USA) with a detection limit of 0.8 μg/ml. Serum gentamicin levels were 1.21 μg/ml at 24 hr following surgery and dropped below detectable limits by five days postoperatively. Gentamicin levels in joint fluid were elevated above serum levels at 30 day post-implantation. In humans, the risk for gentamicin-associated nephrotoxicity is increased when peak serum levels are maintained above 12-14 μg/ml, and trough serum concentrations are greater than 2.0 μg/ml. This data supports the use of AIPMMA beads to provide locally therapeutic drug concentrations with low systemic absorption of a potentially nephrotoxic drug.