Introduction

Tibial plateau leveling osteotomy (TPLO) is a procedure designed to alter the biomechanics of the canine or feline stifle joint to compensate for a torn cranial cruciate ligament (CCL). Although there are multiple surgical techniques for treating CCL disease, the TPLO procedure is the most common among veterinary surgeons.¹ Due to the high volume of cases annually, there is also notably high scrutiny of its infection rate. Previous studies have reported infection rates for TPLO procedures ranging from 3% to 15.8%.^2–5 Some variability in these findings is attributed to inconsistent definitions of postoperative infection. Recently, researchers have begun to rely on the CDC classification for surgical site infection (SSI), with modifications to adapt it to veterinary patients.^2,6 In short, infections can be classified as superficial (skin or subcutaneous tissue), deep, or organ/space incisional infections.⁶ Deep incisional infections in relation to a TPLO procedure often involve infection of the implant, necessitating removal. Organ/space incisional infections involve osteomyelitis and/or joint infection. The average rate of TPLO implant removal following SSI is approximately 2.6% to 3%.^2,7 Surgical site infection rates are higher specifically for TPLO procedures compared to other clean orthopedic procedures.⁸ Speculated factors contributing to this finding include the consistent use of implants during TPLO procedures, predisposing them to implant-associated infections and biofilms, thermal damage to the osteotomy caused by the saw blade, minimal soft tissue coverage over the medial tibia combined with extensive periosteal dissection, and potentially longer surgical or anesthetic times compared to other procedures.⁸

As with any postoperative infection, costs begin to accrue quickly as recheck appointments, cultures, and antibiotics are required. The average cost of a postoperative infection can approach, or surpass, 50% of the cost of the original surgery.⁹ Costs increase further when a second surgery is required to remove an infected implant. In a study by Spencer and Daye,³ the total cost spent on TPLO SSIs in the US was estimated to be $9.6 to $15.9 million per year. There are many strategies to decrease postoperative infections, including iodophor-impregnated adhesive drapes, elastic seal extremity drapes, antimicrobial-impregnated suture, postoperative topical and oral antibiotics, postoperative bandaging, and case selection, but such strategies have varying degrees of success.^10–12

Recently, silver-coated (SC) orthopedic implants have become more widely used in veterinary medicine. Silver has a broad spectrum of antimicrobial activity through numerous mechanisms, including disrupting cell transport, blocking the respiratory chain, and limiting transcription and translation by binding to nucleotides and nucleosides of RNA and DNA.¹³ Research on the antimicrobial properties of silver in human medicine is abundant and generally positive. Silver has been found to have efficacy against biofilm-forming methicillin-resistant Staphylococcus pseudintermedius.¹⁴ Silver-coated orthopedic prostheses were also shown to be nontoxic in canine models, supporting its safe use in veterinary medicine.¹⁵ BioMedtrix recently developed an SC TPLO implant that is projected to decrease periprosthetic infection rates within the first 100 days after surgery. Their silver technology involves silver particles embedded in a polysiloxane (SiOxCy) matrix, which acts as a reservoir for the release of silver ions that are active on the coating surface.¹⁶ Previous papers have shown that this particular silver coating shows no cytotoxicity and exhibits in vitro and ex vivo antimicrobial activity.^17,18 In a research study¹⁴ in which titanium models were coated with a plasma polymer containing silver nanoparticles, the silver coating showed more than 99.98% reduction in the number of CFUs of biofilm-forming methicillin-resistant S pseudintermedius compared to noncoated (NC) specimen.

A recent retrospective study¹⁹ examined postoperative complications and SSI rates in dogs with SC TPLO plates compared to a control group. They found no significant difference in SSI and implant removal rates between the SC plate and the control group. However, limitations existed due to the retrospective nature of the study, including potential underestimation of postoperative complications or infections. They also did not distinguish between superficial, deep, and organ/space SSIs in their results. The goal of this study was to assess the infection rate in SC versus NC TPLO implants in a prospective study design. We hypothesized that the group treated with SC implants would have a lower SSI rate than the group with NC implants.

Methods

Canine TPLO patients at a private practice referral hospital that met the inclusion criteria were enrolled in the study between November 2021 and May 2023. Client consent was obtained prior to surgery.

Dogs included in the study had a CCL tear diagnosed clinically with either positive tibial thrust, positive cranial drawer, or both. All dogs had to have an American Society of Anesthesiologist score of 1 or 2 and ideally weighed between 22.7 and 45 kg, coinciding with recommended weights for either a 3.5-mm short or 3.5-mm standard TPLO plate. All TPLO procedures were performed by a board-certified surgeon, residency-trained veterinary surgeon, or resident under supervision of a board-certified surgeon. Dogs undergoing bilateral TPLO procedures were included, but each stifle was documented individually. Dogs were excluded if they were currently on antibiotics or evidence of pyoderma was detected at the time of surgery, necessitating antibiotics to be prescribed at discharge. Dogs were also excluded if they underwent a concurrent orthopedic procedure or postoperative follow-up was not achieved at any time point.

All dogs were premedicated with injectable analgesia (hydromorphone, 0.1 mg/kg, IV). All dogs were induced with injectable anesthetic agents (midazolam, 0.3 mg/kg, IV; propofol, 2 to 4 mg/kg, IV) and maintained on 1% to 2% isoflurane. The dogs were placed in lateral recumbency, and the affected limb was clipped. A dirty scrub using 2% chlorhexidine and alcohol was performed after clipping. A femoral-sciatic block (0.3 mL/kg bupivacaine and 1 µg/kg dexmedetomidine) with a sterilized needle and nerve stimulator was performed in all dogs undergoing a unilateral TPLO procedure. Dogs undergoing bilateral TPLO procedures received an epidural (0.1 mg/kg morphine and 0.5 mg/kg preservative-free bupivacaine) in lieu of a femoral-sciatic block. A second, clean scrub (single-use examination gloves, 2% chlorhexidine scrub, and alcohol scrub for > 5 minutes) was performed after the patient was moved into the operating room and the leg was suspended. Four quarter drapes secured with towel clamps were used to drape the patient, followed by an extremity orthopedic drape. The distal limb was wrapped with sterile drape material followed by sterile veterinary wrap. Iodine povacrylex (DuraPrep; 3M) was applied to all exposed patient skin and allowed to dry.

Stifle arthroscopy or arthrotomy was performed first to assess the CCL and medial meniscus as well as evaluate the overall joint health. Following arthroscopy, if the medial meniscus was suspected to be torn or if it could not be adequately visualized via arthroscopy, an arthrotomy was then performed. Partial meniscectomies were performed for all meniscal tears. Meniscal releases for intact medial menisci were performed at the discretion of the surgeon. After arthrotomy, the joint was flushed with sterile saline and closed with polydioxanone suture (Ethicon) prior to starting the TPLO. All TPLO plates used were locking plates (BioMedtrix), and use of an SC or NC plate was predetermined and randomized by a coin toss. All TPLO plates were packaged and sterilized via γ irradiation in the same manner. A 3.5-mm short plate was used in patients weighing between 22.7 and 34.5 kg, and a standard plate was used in patients weighing between 34.5 and 45 kg. The surgical incisions were lavaged with sterile saline and closed routinely. The fascia was closed with polydioxanone suture, the subcutaneous layer was closed with poliglecaprone 25 suture (Monocryl; Ethicon), and the intradermal layer was also closed with Monocryl suture in a simple continuous pattern, with suture size determined on the basis of patient weight. Prior to closure of the intradermal layer, up to 5.3 mg/kg of bupivacaine liposome injectable suspension (Nocita) was infiltrated into the peri-incisional tissue using a sterile needle and syringe. No skin sutures or skin staples were placed in any incisions. Prior to leaving the operating room, an adherent bandage dressing (Primapore; Smith & Nephew) was applied over the incision. The bandage remained on until the patient was discharged from the hospital, and a new one was applied if the original bandage fell off or was soiled at any point during hospitalization. Both groups received perioperative antibiotics (cefazolin, 22 mg/kg, IV) 30 to 60 minutes prior to the start of surgery, followed by every 90 minutes intraoperatively.

Postoperatively, both groups continued to receive cefazolin every 8 hours for at least 3 doses or until discharge. Cold compresses were applied to the surgical incision every 8 hours in hospital. Dogs were hospitalized overnight and discharged the following day. Analgesia administered in hospital was determined on the basis of surgeon preference but generally included hydromorphone and an NSAID or gabapentin. Patients were subsequently discharged with the pain medications administered in hospital, but no antibiotics were prescribed or instructed to be given at home. Owners were directed to keep an Elizabethan collar on their dogs at all times for 10 to 14 days postoperatively. Owners were instructed to monitor the incision daily and report any concerns. Follow-up was performed at 2 weeks postoperatively and 8 weeks postoperatively. A final long-term follow-up was performed at the conclusion of the study (> 100 days). At 2 weeks postoperatively, the owners were asked to fill out a questionnaire about the appearance of the incision and the dog’s lameness at home. Owners were encouraged, but not required, to send pictures of the incision via email. The investigator collecting the follow-up data and evaluating the incisions was blinded to the type of plate the dogs had received. For the purposes of analysis, incisional complications were defined as any abnormality of the incision, including redness, heat, swelling, pain, discharge, or dehiscence, regardless of whether treatment was recommended or initiated.

Incisions with mild incisional redness or irritation were monitored, but antibiotics were not initiated unless the incision fit the criteria for an incisional infection. These criteria were based on the CDC guidelines.^3,6,20 Most notably, a superficial SSI had to occur within 30 days and a deep SSI had to occur within 90 days (since placement of an implant). The CDC also specifically states that diagnosis of cellulitis (redness, warmth, and/or swelling) by itself does not classify as a superficial SSI.⁶ Incisional infections were treated with cefpodoxime (5 to 10 mg/kg, PO, q 24 h) or amoxicillin/clavulanic acid (13.75 mg/kg, PO, q 12 h) for 10 to 14 days. Aerobic culture and susceptibility were strongly recommended for any incision with evidence of infection. All infected incisions, if evaluated in person, were cleaned with dilute chlorhexidine and sterile saline.

Results

A total of 65 dogs (73 limbs) were recruited for the study. Eight dogs underwent surgery of both stifles, but only 1 had a bilateral procedure; in the other 7 cases, surgical procedures were performed 2 to 8 months apart. The most common breeds among the study cohort were Pit Bull Terriers (n = 17) and Labrador Retrievers (16), followed by German Shepherd Dogs (6), Golden Retrievers (5), and Boxers (4). There were 4 mixed-breed dogs, 2 each of Rottweiler and Siberian Husky, and 1 each of American Bulldog, Bernese Mountain Dog, Border Collie, Doberman Pinscher, Great Pyrenees, Rhodesian Ridgeback, Schnauzer, and Soft Coated Wheaten Terrier. There were 37 (56.9%) spayed females, 35 (53.8%) neutered males, and 1 (1.5%) intact female. Median (IQR) age across the cohort was 6 years (4 to 7 years), with a range of 1 to 13 years. Median weight was 34.5 kg (31.3 to 38.6 kg), with a range of 21.6 to 45.7 kg.

In surgery, the stifle joint was evaluated via arthroscopy alone in 23 (31.5%) limbs, arthrotomy alone in 34 (46.6%) limbs, and a combination of arthroscopy and arthrotomy in 16 (21.9%) limbs. A partial medial meniscectomy due to a meniscal tear was performed in 24 (32.9%) limbs. A caudal meniscal release following observation of an intact medial meniscus was performed in 10 limbs. Thirty (41.0%) limbs received the 3.5-mm short TPLO plate, and 43 (58.9%) limbs received the 3.5-mm standard plate. The NC TPLO plate was implanted in 35 (47.9%) limbs, and the SC plate was implanted in 38 (52.1%) limbs.

The median (IQR) duration of surgery across the cohort was 85 minutes (64 to 100 minutes). Median (IQR) anesthesia time was 169 minutes (147 to 188 minutes).

Descriptive data for the NC and SC groups are shown in Table 1. There was no difference in age (P = .50) or weight (P = .70) between the groups. There was no difference in sex proportion between groups (P = .28). There was no difference between the SC and NC groups in the proportion of dogs that had arthroscopy alone versus an arthrotomy either with or without arthroscopy (P = 1.0). Compared to the NC group, the SC group had significantly shorter surgery (median 76 minutes vs 94.5 minutes; P = .048) and anesthesia times (median 160 minutes vs 180 minutes; P = .0012).

Follow-up at 2 weeks postoperatively was available for 53 of 73 (72.6%) procedures, and consisted of emailed questionnaires and in-person visits. Twelve owners reported redness, swelling, pain, and/or heat of the incision. However, as none of these incisions had additional signs such as purulent discharge or positive bacterial culture, they did not fulfill the criteria for a superficial SSI. All of these incisions ultimately healed without antimicrobial intervention and without progression to SSI. Five owners reported discharge from the incision and/or partial incisional dehiscence, and all of these dogs were confirmed to have an SSI via in-person rechecks or through primary care records (Supplementary Table S1). Four of these cases had NC implants, and 1 case had an SC implant. A culture was obtained from one of the NC cases and grew β-hemolytic Streptococcus spp and S pseudintermedius, both of which were susceptible to cefpodoxime, a third-generation cephalosporin antibiotic, which was continued for a total of 21 days. The other 4 cases declined culture and were started on empirical cefpodoxime or amoxicillin/clavulanic acid (via the primary veterinarian) for 10 to 14 days, resulting in complete resolution of the infection. Thirteen owners responded no to the question, “Has your pet been wearing the E-collar at all times?” (Supplementary Material S1). Of these 13 cases, 4 of them reported incisional complications (redness or heat of the incision) but none of them developed subsequent incisional infections.

Follow-up at 8 weeks postoperatively was available for 63 of 73 (86.3%) procedures, including in-person visits, primary veterinarian records, and phone calls. Thirty-three patients had in-person visits, with 24 occurring at the study’s institution and 9 occurring with primary or rehabilitation veterinarians. On all in-person examinations, stifles were stable with good range of motion and no pain was documented on palpation of the implants. Radiographs were performed in 24 cases; healing or healed osteotomies were noted in every case, and none showed evidence of osteomyelitis. At 8 weeks, all owners reported that their dogs’ incisions were healed.

Long-term follow-up via phone call, in-person appointments for unrelated medical concerns, or email was available for 38 dogs (45 procedures; 61.6%). Follow-up was obtained an average of 340 days postsurgery (range, 120 to 538 days). Two dogs were deceased due to unrelated causes (heart base mass with pericardial effusion in one case at 98 days postoperatively and unspecified neoplasia in the other at 364 days postoperatively). Of the cases for which follow-up was available, no owners reported long-term problems or complications regarding healing of the surgical limb. One dog in the NC group developed a deep SSI at 92 days postsurgery. A culture was obtained, which grew methicillin-resistant S pseudintermedius. Infections persisted despite multiple courses of antibiotics until ultimate implant removal at 6 months following surgery. Bacterial culture submitted on the implants removed confirmed a persistent methicillin-resistant S pseudintermedius infection. This dog was not previously diagnosed with a superficial SSI, and in fact the owner reported complete uncomplicated healing of the incision at the 8-week follow-up.

Surgical site infections were confirmed in 5 limbs within 30 days of surgery, and an additional infection was confirmed in 1 limb within 100 days, ultimately yielding an infection rate of 8.2% (6 cases) in this study. Four superficial SSIs were documented in the NC group (4/35 [11.4%]) and 1 in the SC group (1/38 [2.6%]). One deep SSI was documented in the NC group (1/35 [2.9%]) and none in the SC group. Thus, a total of 5 infections were documented in the NC group (5/35 [14.3%]) and 1 infection was documented in the SC group (1/38 [2.6%]). This difference was not significant (P = .17). The overall incidence of superficial SSI during the study period was 6.8%. The incidence of deep SSI during the study period was 1.4%. No organ/space SSIs were recorded in this study.

The number of SSIs was too small for meaningful regression analysis; thus, the less stringent outcome of incisional complications at any time point postsurgery was assessed. Incisional complications were reported in 11 of 34 (32.4%) limbs receiving an NC implant for which follow-up was available and in 8 of 34 (23.5%) limbs in the SC group. These proportions were not different between groups (P = .59). Variables considered for inclusion in the multivariable model were implant type (NC vs SC), surgery duration, anesthesia duration, age, weight, sex, surgeon type (diplomate vs other), limb (left vs right), and approach (arthroscopy alone vs arthrotomy with or without arthroscopy). In its final reduced form, only anesthesia duration and age were retained in the model; however, neither reached the level of statistical significance (P = .11 and P = .09, respectively) and their effect was very small (Supplementary Table S2). Post hoc analysis confirmed that the sample size was sufficient to detect a difference in infection rate of 30% between groups.

Discussion

The objective of this study was to examine whether SC TPLO implants have a significant impact on infection rates postoperatively compared to NC implants. We found that both the NC and SC groups had low superficial surgical infection rates (11.4% and 2.6%, respectively) and low deep surgical infection rates (2.9% and 0%, respectively). We therefore reject our hypothesis that the SC implants would yield a significantly lower SSI rate than the NC implants. Although 5 of the 6 SSIs were diagnosed in the NC group, this difference did not reach the level of statistical significance. When all incisional complications were considered, there was also no difference between groups. These findings may be attributed to the sample size, which was underpowered for detection of small differences between groups, as well as the overall low incidence of SSI.

The overall SSI rate in the study cohort of 8.2% is consistent with previous studies in which SSI rates ranged from 3% to 15.8%.^2–4,19 The SC TPLO plate used in this study is designed to decrease the incidence of postoperative infections in the first 100 days after surgery. While by CDC definition, SSI related to procedures such as TPLO are classified as occurring within 90 days of surgery, cases in this study were followed out to at least 100 days because this was the time frame used by the manufacturer when describing the efficacy of their SC TPLO implants. Any evidence of infection between 30 and 100 days was thus considered to be a deep or organ/space SSI. Only 1 case in our study developed a deep (implant) infection, diagnosed 92 days postoperatively, and the implant removal did not occur until 6 months after surgery. As TPLO implant removal has been shown to occur anywhere from 118 days to 16 months after surgery,^2,18,23 it is possible that some long-term infections were missed due to the relatively short duration of this study. However, among the remaining dogs for whom longer-term follow-up was available (average, 340 days postoperatively), no new infections or complications were reported in either group, suggesting that the 100-day follow-up would have been sufficient to detect complications directly related to the surgery or implants. Since our single case of a deep SSI was technically outside of the CDC guidelines of 90 days, it is possible that the CDC criteria may not be effective at defining all SSIs in veterinary medicine and a longer time frame may need to be considered, especially for deep and organ/space SSI in veterinary patients moving forward.

Previous studies have shown that increased risk of SSIs is associated with increased body weight, increased duration of surgery and/or anesthesia, increased age, male sex, and German Shepherd Dogs.^2,11,24,25 Although none of these factors reached statistical significance in our regression analysis, the length of surgery and anesthesia was significantly longer in the NC versus SC group. It is therefore possible that the increased length of the procedure and anesthesia contributed to the incidence of SSI in the NC group. There were no differences in the surgical preparation, procedure, or surgeons involved between the SC and NC groups, so the reason behind the longer surgery and anesthesia times in the NC group is unclear. There was no difference in breed distribution between the SC and NC groups, specifically regarding German Shepherd Dogs. Age, sex, and weight did not differ between the NC and SC groups. No dogs in the study weighed more than 50 kg, which has previously been identified as a risk factor.²⁶

The first follow-up was performed at 2 weeks postoperatively, which is approximately when uncomplicated healing of a skin incision is expected to occur. At this time point, multiple owners reported redness of the incision or some swelling, but these incisions did not fit the criteria for a superficial SSI as described by the CDC guidelines. Indeed, all of these incisions went on to heal uneventfully without additional intervention. A culture was obtained in 2 of the 6 cases with defined SSIs: 1 superficial and 1 deep SSI. Staphylococcus pseudintermedius was present in both cultures that were submitted. These culture results were consistent with previous studies, in which the most common bacteria isolated from TPLO infections was S pseudintermedius.^{1,19,24,27,28}

This study did not detect a difference in SSIs between SC and NC implants, prompting the authors to question the efficacy of SC implants. In theory, the silver coating should provide a protective mechanism against deep and organ/space SSIs due to the local antimicrobial properties of the coating on the implant. The low number of infections within the current study with only 1 deep SSI within the study period did not provide enough data to elucidate any conclusions on this matter. Whether the silver coating has an impact on superficial SSIs remains to be seen. The minimal soft tissue coverage over implants is one of the postulated reasons for an increased SSI rate in TPLO surgeries, and it could be reasoned that an implant’s silver coating impacts more superficial bacteria as a result. There is also no current research measuring the extent that these silver ions diffuse through tissues in a canine TPLO model. Experimental mice models demonstrated that SC implants release silver ions into the surrounding tissues, but the extent of this diffusion was not measured.²⁹ An interesting question would be to determine whether the silver coating reduces the risk of a deep or organ/space SSI following development of a superficial SSI. Studies investigating this would need to be considerably larger, due to the relatively low incidence of these infections following TPLO surgery.

Limitations of this study included its small sample size and short-term follow-up. Post hoc power analysis indicated that to detect a difference in complication rate between groups of 10%, a sample size of 273 limbs/group would be required. Therefore, type II error may underlie the lack of significant differences between the 2 groups despite the numerical differences in SSI incidence. The time frame for this study was set at 100 days on the basis of the purported efficacy of the silver coating on TPLO implants, but the impact of the silver coating might peak beyond 100 days. Additionally, COVID-19 restrictions during the study period limited the number of in-person rechecks that could be performed, so many incisional evaluations were based on owner questionnaires and emailed pictures, which could have falsely increased or decreased our number of documented infections and/or complications. It was not possible to blind the surgeon to the type of plate being used, since the SC plates were gold in color and the NC plates were silver; however, the individual evaluating postoperative follow-up data and images was blinded to implant type, offsetting this possible bias. We were also unable to definitively prove the presence of infection in all cases since not all incisional infections had cultures submitted.

In conclusion, this study showed that while the overall incidence of SSI was low, there was a trend toward increased incidence of incisional infections in limbs with NC TPLO plates compared to SC TPLO plates; however, the small sample size precluded statistical significance of these findings. Surgery and anesthesia times were longer in the NC group than in the SC group, which could have increased the risk of postoperative infection in that group. There was only 1 deep infection and subsequent implant removal within the study, and therefore no conclusions can be drawn on the impact of the silver coating on deep SSIs. The impact of SC TPLO plates on deep and organ/space SSI and implant removal rates should ideally be evaluated by following a larger cohort of cases for 1 to 2 years postoperatively.

Supplementary Materials

Supplementary materials are posted online at the journal website: avmajournals.avma.org.

Disclosures

The authors have nothing to disclose. No AI-assisted technologies were used in the generation of this manuscript.

Funding

Noncoated implants were provided complimentarily by BioMedtrix LLC.