Implant Removal Rate from Infection after Tibial Plateau Leveling Osteotomy in Dogs

Transcription

1 Implant Removal Rate from Infection after Tibial Plateau Leveling Osteotomy in Dogs Alissa D. Gallagher, DVM and W. Daniel Mertens, DVM, Diplomate ACVS Carolina Veterinary Specialists, Charlotte, NC Corresponding Author Alissa Gallagher, DVM, 2225 Township Road, Charlotte, NC Submitted September 2010 Accepted January 2012 DOI: /j X x Objective: To determine implant removal rate associated with infection after tibial plateau leveling osteotomy (TPLO) in dogs and to report antimicrobial susceptibility patterns for isolates. Study Design: Retrospective case series. Animals: Dogs (n = 255; 282 TPLO). Methods: Medical records (April 2006 April 2008) for dogs that had TPLO with 18 month follow-up were reviewed. Dogs that had implant removal with confirmed bacterial isolation from the implant were studied. Cefazolin (22 mg/kg intravenously) was administered before anesthesia induction for TPLO, every 2 hours intraoperatively, and every 6 or 8 hours until the next morning. Antimicrobial susceptibility testing was performed on isolates. Results: Twenty-one (7.4%) of 282 TPLO required implant removal because of infection. Bacterial species isolated were Actinomyces spp. (1), Corynebacterium spp. (1), Enterococcus spp. (3), hemolytic Staphylococcus coagulase negative (2), nonhemolytic Staphylococcus coagulase negative (3), Staphylococcus spp. coagulase positive (7), methicillin oxacillin-resistant Staphylococcus coagulase positive (2), and Serratia marcesens (2). Of the antibiotics that had 10 isolates tested against them, gentamicin had the highest susceptibility rate (94%), followed by tribrissen (71%), and amoxicillin/clavulanic acid (67%). Conclusion: Staphylococcus spp. was reported in 14 of the 21 infections cultured in this study. Based on antimicrobial susceptibility testing, amoxicillin/clavulanic acid would be the best empirical treatment. Tibial plateau leveling osteotomy (TPLO), a surgical procedure believed to stabilize the stifle joint during weight bearing by decreasing cranial tibial thrust 1, is commonly used to manage cranial cruciate ligament (CrCL) rupture in dogs. TPLO does not replace the CrCL or eliminate cranial drawer but allows the caudal cruciate ligament to mimic the function of the CrCL. 1 Surgical procedures for management of CrCL injury have a success rate of 85 90%; however, TPLO is reported to result in quicker return to function. 1 Postoperative complication rates for TPLO vary from %, 2, 3 and include infection, osteomyelitis, implant failure, and seroma. Infection rates range from %. 3, 4 Priddy et al reported an infection rate for TPLO of 13% with incisional infection accounting for 3.1%; draining tracts, 0.5%; septic arthritis, 1.6%; and osteomyelitis, 7.3%. 5 Stauffer et al reported no osteomyelitis whereas Fitzpatrick reported an overall infection rate of 6.6% with 2.6% (n = 26) undergoing implant removal. 3, 6 In that report, 20 cases were diagnosed with infection (culture positive) and 6 dogs had suspected infection (culture negative). 3 All other studies investigating complications after TPLO reported similar rates. 2, 5 We are unaware of a study reporting implant removal rate because of infection based on positive microbial culture. Those considered initially infected may have been infected or may have had a superficial infection. We defined infected implants as those that were culture positive. Implant-associated infection rates have been reported to be 5% 7 and % 4 in human and veterinary medicine, respectively. Whereas many TPLO complications can be managed conservatively, infection typically requires surgical removal of the implant. 8, 9 Ultimately, this increases the owner s financial and emotional burden and increases risk of patient morbidity. Repeat surgery of 7.5% has been reported for lateral fabellotibial suture. 10 Our purpose was to determine the rate of implant removal associated with infection after TPLO. A secondary objective was to make recommendations about prophylactic and empirical antibiotic selection based on culture and susceptibility testing performed at the time of implant removal. This information is important clinically when a diagnosis of implant-associated infection is suspected before union of the tibial osteotomy. Such cases are generally managed by empirical treatment of the suspected infection until satisfactory bone healing occurs, then the TPLO plate may be Veterinary Surgery 41 (2012) C Copyright 2012 by The American College of Veterinary Surgeons 705

2 Implant Removal after TPLO Gallagher and Mertens removed and cultured, and definitive antimicrobial therapy administered. Unfortunately, microbial culture and susceptibility testing results derived from draining tracts do not accurately reflect deep cultures obtained from the implant. According to Akinyoola et al, the reported susceptibility, specificity, and predictive value of draining tracts compared with intraoperative bone cultures of chronic osteomyelitis are 50%, 20%, and 50%, respectively. 11 Significant risk occurs to the patient if inappropriate empirical antibiotic therapy is selected; osteomyelitis with delayed or nonunion may result. 1 In severe cases, time lost because of inappropriate empirical antibiotic selection could lead to severe complications such as amputation or necessity for another major surgery. 12 MATERIALS AND METHODS Medical records (April 2006 April 2008) of dogs that had TPLO were retrieved. Dogs were included if there was 18 months follow-up after TPLO. Retrieved data included breed, sex, age, weight, affected limb, time from surgery to infection, culture and susceptibility testing results, empirical treatment selected, susceptibility to empirical treatment, type of infection noted, miscellaneous information noted that may have contributed to infection, and the date and time of the procedure. Data Classification Dogs were assigned to 1 of 2 groups: no implant removal or implant removal because of infection. Dogs were classified as requiring implant removal for infection if there was a positive microbial culture from the implant, which provides definitive diagnosis of an implant-associated infection. If a dog was administered antibiotics because the incision appeared abnormal (erythematous, edematous, tenderness) but did not require implant removal, it is difficult to state whether the incision was infected or inflamed. Thus, if a dog was administered antibiotics at any time after TPLO for a suspected incisional infection and treated empirically with antibiotics, but did not ultimately require plate removal, they were not counted as requiring plate removal associated with implant infection. If implants were removed because of a suspected infection but this was not confirmed by a positive microbial culture, the case was not classified as requiring implant removal because of infection. Dogs were suspected of implant-associated infection for various reasons, the 2 most common being draining tracts associated with the surgical site or a return of lameness after a seemingly normal recovery. Clinical findings often included lameness with no other identifiable cause (no meniscal pathology suspected) and pain on palpation of the bone plate. Recheck radiographs were obtained because of these clinical signs; however, zones of lucency or other radiographic signs of infection typically do not appear until 2 weeks after onset of infection. 13 Subtle radiographic signs of infection are difficult to interpret, so radiographs alone are rarely diagnostic for infection. Slight periosteal reaction could be a sign of infection but may also be because of periosteal elevation of the proximomedial aspect of the tibia during TPLO. Early subperiosteal callus formation may also appear similar to a periosteal reaction caused by osteomyelitis. Lucency around the screws may be a sign of infection; however, with digital radiographs, the Uberschwinger artifact (lucency around metallic implants created artifactually by imaging processing) can make interpretation of this finding difficult. 14 Correct interpretation of radiographs improves when findings are correlated with other clinical information. TPLO TPLO were performed by 1 of 2 board-certified surgeons in an operating suite dedicated to clean surgical procedures. Cefazolin (22 mg/kg intravenously [IV]) was administered between induction and surgery, every 2 hours throughout the procedure, and every 6 8 hours postoperatively until the next morning. Whereas the surgical procedure is relatively short, time under anesthesia for TPLO may be extended because of preoperative radiographs, surgical site preparation, administration of an epidural, postoperative radiographs, and bandaging before recovering from anesthesia, and in our clinic are typically 3 hours for TPLO. We use a medial approach (skin incision from the distal 4th of the femur to the proximal 3rd of the tibia) and stifle arthrotomy to remove the CrCL remnants and inspect the meniscus. The medial meniscus was debrided appropriately if damaged. Depending on surgeon preference, meniscal release may or may not have been performed. A Slocum TPLO saw blade was used to perform the tibial osteotomy. The proximal tibia was rotated based on a preoperative tibial plateau slope calculated from radiographs using standard guidelines and stabilized using a TPLO plate (Securos, Fiskdale, MA), and either Securos or New Generation (Glen Rock, NJ) screws were applied to the osteotomy. The incision was closed in layers with stainless steel staples used for skin apposition. A modified Robert Jones bandage was placed until the next morning. Dogs were hospitalized overnight, administered IV analgesia and cefazolin (every 6 or 8 hours depending on surgeon preference), and discharged the following day. Staple removal was scheduled days postoperatively and radiographs were performed 8 weeks postoperatively to assess bone healing. Dogs that had superficial dermatitis at surgery were administered cephalexin ( 20 mg/kg twice daily for 7 days) postoperatively. Other postoperative medications (nonsteroidal anti-inflammatory drugs, gabapentin, and/or tramadol) were prescribed for 1 week, then as needed, for pain and discomfort as perceived by the owner. Peri- and Postoperative Incisional Complications Perioperative incisional inflammation/infection (superficial dermatitis, drainage, seroma, or possible suture 706 Veterinary Surgery 41 (2012) C Copyright 2012 by The American College of Veterinary Surgeons

3 Gallagher and Mertens Implant Removal after TPLO reaction) was treated empirically. If drainage occurred, empirical treatment or microbial culture was performed at surgeon discretion. If empirical treatment was used, duration was 7 14 days depending on surgeon. Those that resolved were not included in this study because they were classified as superficial not implant-associated infection. Dogs with persistent lameness, pain on palpation of the plate, or lack of resolution of drainage if present were treated empirically or microbial culture was performed if there was drainage. Dogs with persistent lameness, or pain on palpation of the plate, were treated empirically with cephalexin, clavamox, or enrofloxacin. If drainage was persistent, microbial culture was performed and the dog treated according to biogram results. Any treatment regimen started was continued until implant removal (mean, 99 days; range, days). Plate Removal and Bacterial Culture and Susceptibility Testing Perioperative antibiotics were withheld until after implant removal. An incision was made on the medial aspect of the tibia directly over the bone plate and the screws and bone plate were removed. Microbial cultures were obtained by submitting a screw for aerobic and anaerobic culture and susceptibility testing. Data Analysis All Staphylococcus coagulase positive and negative organisms were tested against an antibiogram containing augmentin, ampicillin, ceftiofur, cephalothin, clindamycin, enrofloxacin, erythromycin, gentamicin, oxacillin, penicillin G, tetracycline, tribrissen, and chloramphenicol or vancomycin. Staphylococcus resistant to methicillin/oxacillin were tested against augmentin, ampicillin, ceftiofur, cephalothin, chloramphenicol, clindamycin, enrofloxacin, erythromycin, gentamicin, oxacillin, penicillin G, tetracycline, tribrissen, and rifampin. Enterococcus was tested against an antibiogram containing ampicillin, tetracycline, augmentin, and either chloramphenicol and penicillin G or enrofloxacin and tribrissen. Serratia marcesens isolated from thio broth was tested against an antibiogram containing augmentin, ampicillin, carbenicillin, cephalothin, ciprofloxacin, enrofloxacin, gentamicin, tetracycline, tricarcillin, tobramycin, tribrissen and 1 antibiogram also contained amikacin, ceftazidime, ceftiofur, chloramphenicol, and piperacillin. An isolate of S. marcesens from agar was tested against cefazolin, cefixime, cefotaxime, ceftriaxone, cefuroxime, difloxacin, imipenem, marbofloxacin, orbifloxacin, ofloxacin, and cefpodoxime. Actinomyces spp. were susceptible to penicillin, erythromycin, clindamycin, ampicillin, doxycycline, and cephalosporins. Corynebacterium spp. aresusceptibleto penicillin. RESULTS A total of 282 (159 left, 123 right) TPLO were performed on 255 dogs. Five dogs that had bilateral staged procedures were infected unilaterally and 1 dog was infected bilaterally. The most common breeds were mixed breed (n = 77) and Labrador retriever (66), followed by Golden retriever (17), Rottweiler (10), and Mastiff (10). Twenty-seven other breeds were also represented. Mean ± SD age at surgery was 5.7 ± 2.6 years (range, 1 13 years). Mean weight was 35.7 ± 11.5 kg (range, kg).gender distribution was 131 spayed females (51.3%), 101 neutered males (39.6%), 6 sexually intact females (2.3%), and 15 sexually intact males (6.7%). Forty four (17.3%) of the 255 dogs were diagnosed with dermatitis, drainage, seroma, or possible suture reaction postoperatively and treated empirically with cephalexin ( 20mg/kg). Two of these dogs had bilateral TPLOs and subsequently dermatitis diagnosed at both procedures. Of 282 TPLO, 24 (8.5%) had implant removal because of continued lameness and suspected infection. Twenty-one (7.4%) were diagnosed with an implant-associated infection based on positive culture results. Three paired cultures (drainage and implant) were also available for analysis. Coagulase positive Staphylococcus spp. was isolated from the drainage in all 3 dogs; however there was no isolate from the implant in 1 dog and a different species of Staphylococcus spp. and Enterococcus spp. were isolated from the implants in the other 2 dogs. Three implants were culture negative and these dogs were being administered either clindamycin (2 dogs) or ciprofloxacin (1) until and after implant removal. One of these dogs with a bilateral infection had coagulase positive Staphylococcus spp. isolated but each isolate had a different susceptibility pattern. Eight causative organisms were identified from microbial culture of implants. Of 21 positive implant culture and susceptibility results, 14 isolates (66.7%) were Staphylococcus spp.: nonhemolytic coagulase negative Staphylococcus (n = 3, 21.4%); coagulase positive Staphylococcus spp. (n = 7, 50 %); hemolytic coagulase negative Staphylococcus (n = 2, 14.3); and methicillin oxacillin-resistant coagulase positive Staphylococcus (n = 2, 14.3%). Other organisms isolated included Enterococcus spp. (n = 3, 21.4%), and S. marcesens (n = 2, 14.2%; Fig 1). Corynebacterium spp. (n = 1) and Actinomyces (n = 1) were also isolated, neither of which had susceptibilities reported, but these are typically susceptible to penicillin and penicillin, erythromycin, clindamycin, ampicillin, doxycycline, and cephalosporins, respectively. Not all organisms were tested against all antibiotics. With such a small sample size (21 cultures), we discarded results from antibiotics that had <10 isolates tested. Of the antibiotics that had 10 isolates tested against them, gentamicin had the highest susceptibility rate at 94% (15/16), followed by tribrissen at 71% (10/14), amoxicillin/clavulanic acid at 67% (12/18), penicillin G at 64% Veterinary Surgery 41 (2012) C Copyright 2012 by The American College of Veterinary Surgeons 707

4 Implant Removal after TPLO Gallagher and Mertens Figure 1 Culture and susceptibility results (susceptible = black, intermediate = gray, resistant = white). 708 Veterinary Surgery 41 (2012) C Copyright 2012 by The American College of Veterinary Surgeons

5 Gallagher and Mertens Implant Removal after TPLO Table 1 Susceptibility Rates of Bacteria to Antibiotics Tested on >10 Isolates Antimicrobial # Susceptible Total % Susceptible Ampicillin % Augmentin % Ceftiofur % Cephalothin % Chloramphenicol % Clindamycin % Enrofloxacin % Gentamicin % Oxacillin % Penicillin G % Tetracycline % Tribrissen % (6/16), oxacillin 64% (9/14), penicillin G at 64% (6/16), chloramphenicol at 60% (6/10), cephalothin at 54% (7/13), tetracycline at 53% (9/17), clindamycin at 50% (5/10), ceftiofur at 40% (6/15), enrofloxacin at 31% (8/18), and ampicillin at 31% (6/19; Table 1). DISCUSSION Our results show that 8.5% (n = 24) of dogs that had TPLO had later surgery for implant removal. Of these, 7.4% (21) had positive cultures and were classified as implant removal because of infection. Three dogs (1.1%) were not classified as having implant-associated infection because the implants were culture negative. These 3 dogs were being administered antibiotics based on culture and susceptibility testing of percutaneous drainage and interestingly, isolates from the implant and drainage were different. This finding is supported by Darouiche s observation that percutaneous drainage culture has a specificity of 50% to diagnose implant-associated infections. 15 Although there is a low specificity, our cases were administered antibiotics based on the biogram for drainage for 4 6 weeks before plate removal. Lameness continued in all 3 dogs and therefore the implants were removed, so it is possible that the organism was susceptible to the antibiotic and hence the negative implant culture or there was another underlying cause. In 2 of these dogs, the cause of the ongoing lameness was never resolved and in 1 dog, contracture and fibrosis of the cranial tibial muscle was eventually diagnosed as the cause of the lameness. Our findings of 7.4% implant-associated infection correlates well with the 7.3% of TPLO cases that developed osteomyelitis reported by Priddy et al 5 ; however, our finding is 2.7 times higher than that reported by Fitzpatrick et al. 3 Overall infection rates for all surgeries are reported to be %. 3, 4 In people, 5% of orthopedic implants become infected. 7 Whereas our infection rate is within the range of previously reported results, 7.4% seems high for a clean procedure with healthy surrounding tissues and a short surgical procedure ( 1 hour in our hospital). A potential contributing reason could be limited soft tissue coverage where the bone plate is placed. If there is a mild incisional infection, the bone plate may become contaminated resulting in an inability of the body to clear an otherwise mild infection. Thermal bone necrosis caused by the saw blade may raise the risk of infection relative to a fracture. Length of anesthesia has been well documented to correlate with postoperative infection rates. 16 We use propofol for anesthetic induction for TPLO and propofol has been shown to increase the risk of superficial and deep surgical infections. 17 Use of stainless steel staples is also associated with a higher 18, 19 infection rate. Fitzpatrick et al, suggested that infection rates were lower in dogs that were administered postoperative cephalexin for 14 days. 3 Most organisms identified in their biograms were susceptible to cephalexin 3 ;however,mostof our isolates were not susceptible to cephalexin. Although some of our dogs were treated for perioperative conditions, this should not have affected our results. Further studies are needed to determine if postoperative cephalexin would decrease our rate of implant removal. An implant-associated infection is a major complication that in general requires device removal for resolution, because of formation of a biofilm on the implant. Biofilm development begins with bacterial adherence via host-tissue ligands present on the implant such as fibronectin and fibrinogen. Depletion of metabolic substrates and waste product accumulation within the biofilm combine to cause bacteria to enter a slow- or nongrowing state, whereby they become up to 1000 times more resistant to antimicrobial agents. Once microbes attach to an implant, host cells have difficulty displacing them 8, 9 ;however,notallbacteriaare capable of producing biofilms. Bacteria most commonly reported to cause implantassociated infections are Staphylococcus spp. and Streptococcus spp. 1, 2, 4, 5, 20 with Staphylococcus spp. accounting for 60% of these infections. 20 Our results are consistent with these findings. Staphylococcus spp. are thought to be commonly associated with implant infections because they are resident skin flora. Staphylococcus spp. isolates are often resistant to many antibiotics with up to 50% of strains being resistant to methicillin. 21 Strains are also been reported to be resistant to vancomycin. 21 Staphylococcus spp. are capable of forming biofilms. 20 In veterinary practice, cephalexin is considered a good choice for empiric treatment of orthopedic surgical infections because of low cost, oral administration, good penetration into bone, and a broad spectrum of activity that generally includes effectiveness against Staphylococcus spp. and Streptococcus spp. We acknowledge that our culture and susceptibility findings reflect our hospital environment and recommendations may not be fully applicable to other hospitals where flora isolated may be different. Most (14/21) isolates were Staphylococcus or Streptococcus and only 7/14 (50%) isolates were susceptible to either cefazolin or cephalothin. Other antibiotics that are often considered for similar reasons are amoxicillin/clavulanic acid and fluoroquinolones. Veterinary Surgery 41 (2012) C Copyright 2012 by The American College of Veterinary Surgeons 709

6 Implant Removal after TPLO Gallagher and Mertens Amoxicillin/clavulanic acid is quite expensive, especially in larger animals. Fluoroquinolones are not recommended against methicillin-resistant Staphylococcus aureus (MRSA) even when initial culture results indicate susceptibility because MRSA invariably develops resistance to fluoroquinolones during the course of treatment. Fluoroquinolones may be used against other Staphylococcus species. Clindamycin and chloramphenicol may also be considered good choices because of their high penetration into bone and the fact that they are often effective against certain strains of MRSA. Use of chloramphenicol is generally avoided because of its ability to cause aplastic anemia in people; however, chloramphenicol has a very broad spectrum of activity against Gram-positive bacteria (including most strains of MRSA), Gram-negative bacteria, and anaerobes. It is not active against Pseudomonas aeruginosa or Enterobacter species. We did not isolate Pseudomonas;however,Enterobacter spp. accounted for 3 of the 14 cases. Based on our culture and susceptibility results, use of amoxicillin/clavulanic acid (67% of isolates susceptible) as the first choice in empirical treatment of suspected TPLOassociated infections is recommended over cephalexin (only 50% of isolates susceptible), although cost remains a consideration. Only 31% (8/18) of isolates were susceptible to enrofloxacin making it a poor first choice for empirical treatment. Chloramphenicol was only tested against 10 isolates of which 5 were susceptible and clindamycin was only tested against 10 isolates of which 5 were susceptible; the small numbers of isolates tested against these antibiotics would make recommendations regarding their use speculative. It is notable that a 67% isolate susceptibility rate exists to even the best choice, amoxicillin/clavulanic acid. Gentamicin and tribrissen have higher susceptibility; however, because of side effects, gentamicin is generally not considered an empirical choice antibiotic. Tribrissen is not an efficacious antibiotic when treating dermatitis. 8 Whereas changing antibiotics can lead to development of resistance, one should be prepared to monitor these cases closely and be prepared to make changes if clinical response to the first choice is poor. In human medicine, there are established recommendations for treatment of implant-associated infections based on culture results. Rifampicin has good efficacy against slow-growing and adherent Staphylococcus spp. infection and as such is the cornerstone of most therapeutic regimens. Because of intratherapeutic development of resistance when rifampicin is used as a sole agent, it is always combined with another antimicrobial. In methicillin susceptible Staphylococcus infections, the oral β-lactamases nafcillin or (flu)cloxacillin are recommended for 2 weeks followed by rifampicin plus ciprofloxacin. For methicillin resistance (excluding S. aureus), rifampicin is combined with vancomycin for 2 weeks followed by rifampicin plus ciprofloxacin. In cases of MRSA, fluoroquinolones are never recommended as resistance invariably develops during treatment. Some surgeons believe any implant-associated infection should be treated presumptively for MRSA in the absence of guiding culture results. Only 2 of our 14 isolates were methicillin/oxacillin-resistant Staphylococcus spp. making this recommendation difficult to follow. Prophylactic therapy against MRSA is only recommended at institutions with high MRSA infections rates. Most of our isolates were pathogenic coagulase positive Staphylococcus spp. Six of 7 of the coagulase positive Staphylococcus spp. we isolated were susceptible to amoxicillin/clavulanic acid. We found that 8.5% of our dogs had surgery for treatment of suspected implant-associated infection, of which 7.4% were culture positive. A 2nd surgery increases patient morbidity and owner cost. If drainage is present, a percutaneous culture is often obtained; however, the major limitation of this is low specificity for deep infections. If drainage is not present or empirical therapy is instituted, it should be used until sufficient osteotomy healing occurs to allow implant removal. Identifying the commonly isolated organisms and their resistance patterns can help guide empirical treatment until plate removal. With the antimicrobial susceptibility results we obtained, amoxicillin/clavulanic acid would be the best empirical treatment because 67% of the cultured organisms were susceptible. REFERENCES 1. Dejardin LM: Tibial plateau leveling osteotomy, in Slatter D (ed): Textbook of small animal surgery, Vol. 2 (ed 3). Philadelphia, PA, Saunders, 2002, pp Pacchiana PD, Morris E, Dillings SL, et al: Surgical and postoperative complications associated with tibial plateau leveling osteotomy in dogs with cranial cruciate ligament rupture: 397 cases ( ). JAmVetMedAssoc 2003;222: Fitzpatrick N, Solano MA: Predictive variables for complications after TPLO with stifle inspection by arthrotomy in 1000 consecutive dogs. Vet Surg 2010;39: Wosar M: Preventing infection in orthopedic surgery. NAVC Proceedings North American Veterinary Conference, Orlando, FL, January 2007, pp Available at Stauffer KD, Tuttle TA, Elkins AD, et al: Complications associated with 696 tibial plateau leveling osteotomies ( ). JAmAnimHospAssoc2006;42: Priddy NH, Tomlinson JL, Dodam JR, et al: Complications with and owner assessment of the outcome of tibial plateau leveling osteotomy for treatment of cranial cruciate ligament rupture in dogs: 193 cases ( ). JAmVetMedAssoc 2003;222: Trampuz A, Zimmerli W: Diagnosis and treatment of infections associated with fracture-fixation devices. Injury 2006;37: Hoyle BD, Costerton JW: Bacterial resistance to antibiotics: the role of biofilms. Prog Drug Res 1991;37: Veterinary Surgery 41 (2012) C Copyright 2012 by The American College of Veterinary Surgeons

7 Gallagher and Mertens Implant Removal after TPLO 9. von Eiff, Proctor RA, Peters G: Coagulase-negative staphylococci. Pathogens have major role in nosocomial infections. Postgrad Med 2001;110:63 70, Casale SA, McCarthy RJ: Complications associated with lateral fabellotibial suture surgery for cranial cruciate ligament injury in dogs: 363 cases ( ). JAmVet Med Assoc 2009;234: Akinyoola AL, Adegbehingbe OO, Aboderin AO, et al: Therapeutic decision in chronic osteomyelitis: sinus track culture versus intraoperative bone culture. Arch Orthop Trauma Surg 2009;129: Bubenik LJ, Smith MM: Orthopedic infections, in Slatter D (ed): Textbook of small animal surgery, Vol. 2 (ed 3). Philadelphia, PA, Saunders, 2002, pp Carek PJ, Dickerson LM, Sack JL: Diagnosis and management of osteomyelitis. Am Fam Physician 2001;63: Caine A: Practical approach to digital radiography. In Practice 2009;31: Darouiche RO: Treatment of infections associated with surgical implants. New Engl J Med 2004;350: Eugster S, Schawalder P, Gaschen F, et al: A prospective study of postoperative surgical site infections in dogs and cats. Vet Surg 2004;33: Nichols RL: Preventing surgical site infections: a surgeon s perspective. Emerg Infect Dis 2001;7: Frey TN, Hoelzler MG, Scavelli TD, et al: Risk factors for surgical site infection-inflammation in dogs undergoing surgery for rupture of the cranial cruciate ligament: 902 cases ( ). JAmVetMedAssoc2010;236: Smith TO, Sexton D, Mann C, et al: Sutures versus staples for skin closure in orthopaedic surgery: meta-analysis. BMJ 2010;340:c Soontornvipart K, Nečas A, Dvořák M: Effects of metallic implant on the risk of bacterial osteomyelitis in small animals. Acta Vet Brno 2003;72: Drew RH. Emerging options for treatment of invasive, multidrug-resistant staphylococcus aureus infections. Pharmacotherapy 2007;27: Veterinary Surgery 41 (2012) C Copyright 2012 by The American College of Veterinary Surgeons 711