Introduction

Pancarpal arthrodesis (PCA) is a salvage procedure that involves fusion of the antebrachiocarpal, middle carpal, and carpometacarpal joints.¹ Immobilization of the carpus by osseous union is indicated in dogs with carpal hyperextension, shearing injuries, luxations, irreparable articular fractures, immune-mediated arthritis, and degenerative joint disease not responsive to conservative treatment.²

To achieve bone fusion, rigid stabilization of the joint is necessary. This can be achieved using bone plates² or external skeletal fixation.^3,4 The use of different types of plates has been reported, including dynamic compression plates,⁵ hybrid dynamic compression plates,^6–10 veterinary cuttable plates,¹¹ castless plates,^8,12 stepped hybrid plates,¹³ and locking plates.^14–16

Bone plates can be applied in the cranial/dorsal,² medial,¹⁷ or palmar/caudal surfaces.¹⁸ Percutaneous techniques have also been reported.¹⁹ The compression side of the carpus is the dorsal aspect, and the tension side is the palmar aspect, meaning that palmar/caudal application is a biomechanically superior position to place the plate.¹⁸ The disadvantage of the palmar technique is the challenging surgical approach through the soft tissues.¹⁸ Medial plating is more resistant to bending forces because it is edge-loaded and therefore has an increased area moment of inertia.¹⁷ For this same reason, contouring of the plate (if necessary) in the axial plane of the bone is more challenging. Reconstruction plates can be used to overcome the challenge of contouring; however, it comes at a cost of reduced plate stiffness.

Various augmentations to PCA with cranial/dorsal plates have been described with the aim of reducing implant failure due to cyclic fatigue. A hybrid plate combined with crossed Kirschner wires was described in 2008 by Arnott et al. in an “in vitro” study.²⁰ In 2019, Tuan et al. described the use of locking compression plates with crossed cortical screws.¹⁶ This study did not report a lower complication rate when compared with other studies using single plates. This use of orthogonal bone plates has been described for feline pantarsal arthrodesis with good outcomes.²¹ In the canine carpus, orthogonal plates have been successfully used for stabilization in limb-sparing surgery²² and in PCA after highly comminuted articular fractures.²³

In the past, some studies recommended the use of rigid external coaptation for several weeks,^{1,2,5–8,11,17,19} despite the high complication rates reported with cast application.^2,5 More recently, a limited number of studies showed good clinical outcomes in dogs with single-plated PCA without the need of rigid external coaptation.^8,10,16 However, these studies did have prolonged nonrigid external coaptation.

Orthogonal plating has been reported to reduce fracture gap strain for orthopedic procedures.²⁴ It has also been shown to provide greater construct stiffness and a higher failure load when compared with single plating and plate-rod constructs.²⁴ The objective of this study is to report the short-term, medium-term, and long-term outcomes and complications in dogs treated with orthogonal plated PCA without prolonged rigid or nonrigid external coaptation.

Materials and Methods

The surgical records of a single veterinary referral hospital (Southern Counties Veterinary Specialists) from 2015 to 2022 were searched for cases of canine PCA with a minimum follow-up period of 8 mo. Dogs included underwent unilateral PCA using orthogonal plating without the use of postoperative rigid external coaptation. All the surgeries were performed by a board-certified surgeon, or a small animal surgical resident under their direct supervision. Data collected included breed, age, body weight, sex, indication for arthrodesis, implants used, intraoperative and postoperative complications, and clinical outcome. Postoperative radiographs were reviewed to assess radiographic signs of bone healing, screw loosening, implant breakage, and other potential complications.

Long-term follow-up (as defined by Cook et al.²⁵) was assessed by owner questionnaire (Canine Brief Pain Inventory [CBPI]) by telephone and/or email communication to assess the functional outcome of the patient. This retrospective study was approved by the Royal College of Veterinary Surgeons Ethics Review Panel.

Postoperative Management

Postoperative orthogonal radiographs of the operated leg were obtained to assess implant positioning and limb alignment (Figure 1). A modified Robert-Jones dressing was placed to control swelling. The duration of the dressing ranged from 24 hr to 7 days, depending on surgeon preference, with an average of 3 days. All patients were prescribed with postoperative analgesia, and the majority received postoperative antibiotics.

View Figure

• Download as Powerpoint
• Download Figure

Postoperative management involved crate or small room confinement, with just short lead walks for toilet breaks for the first 2 wk. Thereafter, short (5–10 min) periods of lead walks three to four times daily were permitted, maintaining confinement the remainder of the time. This persisted until the next follow-up (usually 8 wk postoperatively).

Short-Term Follow-Up

Clinical and radiographic reassessment was performed at 4–12 wk postoperatively. Some dogs required a second or third clinical evaluation and radiographic reassessment at the clinicians’ discretion. Lameness was categorized and scored by the attending clinician. Limb function was categorized as none or mechanical lameness (0/5), mild (1/5), mild to moderate (2/5), moderate (3/5), moderate to severe (4/5), or severe (5/5). Mechanical lameness was defined as pain-free restricted limb movement.²⁶ Definitions of lameness: mild was defined as subtle lameness only observed at trot; mild to moderate was defined as mild lameness present at walk that was made worse at trot; moderate lameness was defined as obvious lameness at walk and trot; moderate to severe lameness was defined as obvious lameness at walk and trot with intermittent non–weight bearing; and severe lameness was defined as minimal or non–weight-bearing lameness most or all the time.²⁶

The postoperative radiographs were evaluated by the clinician at the time of the clinical re-examination. Radiographs were reviewed retrospectively by the authors, and an arthrodesis scale was given to each joint (radiocarpal, intercarpal, and carpometacarpal). The degree of arthrodesis was described as “none” (no mineralized tissue was visible in the joint spaces), “early” (cancellous bone was bridging the joint but with a space still visible), “moderate” (cancellous bone was bridging the joint and no space was visible, but the subchondral bone plate was present), or “advanced” (there was solid fusion of the bone, and the subchondral bone plate was not visible).²⁷

Complications were recorded and classed as minor, major, and catastrophic (as defined by Cook et al.²⁵). Minor complications were defined as complications that could be managed without additional surgery or minimal medical treatment. Major complications were defined as a complication necessitating revision surgery or medical treatment to resolve a problem as a standard of care.²⁵ Catastrophic complications were defined as causing permanent unacceptable function, limb amputation, euthanasia, or death.²⁵ Surgical site infections (SSIs) were considered a major complication. To provide a more detailed criteria for SSIs, they were further classified as deep SSI (infection of fascia, muscle layers, or deeper tissues) or as superficial SSI (infection involving the skin and subcutaneous tissues).²⁸

If no complications were identified and the clinical function was optimal, indications of gradual increase in activity over the following 8 wk were provided to the owner, as was the recommendation of physiotherapy. Further clinical re-examinations were requested for dogs with suboptimal limb function or delayed healing of the arthrodesis.

Long-Term Follow-Up

Long-term follow-up was assessed by the CBPI questionnaire. Owners were contacted by telephone and/or email; after verbal or written consent for the use of data collected, they were asked to complete the CBPI.

Results

Twenty-one dogs met the inclusion criteria; all cases had unilateral orthogonal plated PCA performed. There were 11 neutered females, 5 castrated males, 3 entire males, and 2 entire females. Breeds included 3 English springer spaniels, 3 collie crossbreed dogs, 2 medium-large crossbreed dogs, 2 border collies, 2 Siberian huskies, 1 golden retriever, 1 rottweiler, 1 Cavalier King Charles spaniel, 1 Hungarian vizsla, 1 Weimaraner, 1 Bernese mountain dog, 1 German shepherd dog, 1 shar pei, and 1 miniature schnauzer. The mean weight of all dogs was 23.2 kg (range 10.5–48.5). The mean age at the time of procedure was 3.9 yr (range 1.0–11.0).

Fourteen dogs presented with right carpal injuries and seven with left carpal injuries. The most common indication for PCA was hyperextension injury (13/21; 61.9%), followed by intermedioradial carpal bone fracture (5/21; 23.8%), fourth carpal bone fracture (1/21; 4.8%), degenerative joint disease (1/21; 4.8%), and antebrachial limb deformity (1/21; 4.8%). The plates used for the dorsal aspect were various sizes (depending on patient weight) of hybrid pancarpal arthrodesis plates^g or locking compression plate pancarpal arthrodesis plates^h. For the medial aspect, plates used were different sizes of veterinary cuttable platesⁱ, one 2.7 mm locking polyaxial plate^j, and one 2.7 mm dynamic compression plate (DCP)^k.

Sixteen (76.2%) patients were prescribed with postoperative oral antibiotics, with a mean duration of 7.06 days (range 5–10 days). From these patients, 13 received cephalexin^l and 3 received amoxicillin/clavulanic acid^m. Three patients received IV cefuroxime^f postoperatively for 24 hours, and two patients received IV cefuroxime^f only intraoperatively.

Clinical and Radiographic Follow-Up

The initial clinical re-examination was performed at a mean of 7.2 wk (range 4–12) after surgery. Four (19.1%) dogs had either no lameness or a mechanical lameness, 12 (57.1%) dogs presented with mild lameness, 3 (14.3%) dogs had mild to moderate lameness, and 2 (9.5%) dogs (cases 8 and 14) had severe lameness at the time of re-examination.

Postoperative radiographic evaluation was performed at the time of the initial re-examination. Ten patients had more than one re-examination, and seven of them had more than one radiographic evaluation. The patients that had more than one radiographic evaluation had either a minor or a major complication. The arthrodesis healing scales are represented in Table 1.

TABLE 1 Radiographic Arthrodesis Healing Scale, Complications, and CBPI Scores in Dogs with Orthogonal Plated Pancarpal Arthrodesis

View Fullscreen

Discussion

To the authors’ knowledge, this is the first retrospective study documenting the use of orthogonal plating for PCA in dogs. The results suggest that orthogonal plating can be used for PCA without postoperative rigid coaptation and without prolonged nonrigid dressings. It also showed good long-term outcomes in terms of limb function and comfort, determined by the CBPI scores.

All dogs in this study were adults (mean 3.9 yr, range 1.0–11.0), medium to large dogs were overrepresented (mean 23.2 kg, range 10.5–48.5), and the most common indication for PCA in the present study was carpal hyperextension injury (13/21; 61.9%). These findings are similar to those described in previous publications.^10,16

In the present study, the minor and major complication rate was 4.8% and 33.3%, respectively. These results contrast with the publication by Ramirez and Macias,¹⁰ where they reported a lower major complication rate (17%) but greater minor complication rate (23%). The difference in complication rate is likely due to the different complication classification system, where they classified SSI resolved with oral antibiotics (without the need of implant explantation) as minor complications. If the authors had used the same classification system instead of that described by Cook et al.,²⁵ the major and minor complication rates would have been 14.2% (3/21) and 23.8% (5/21) respectively, which is similar to the Ramirez and Macias study.¹⁰

A study in dogs following PCA using hybrid dynamic compression plates or CastLess plates reported 5.6% and 8.6% for minor complications, along with 28% and 32% for major complications, respectively.⁸ This study has a similar complication rate to the current study, and they classified SSI as a major complication, regardless of the type (superficial or deep), which may explain why the major complication rate in the study by Ramirez and Macias¹⁰ may appear subjectively lower when comparison is made. Furthermore, the high major complication rate in the aforementioned study⁸ could also be explained because of its retrospective nature and that, therefore, not all cases were confirmed with bacteriologic testing. This is a scenario that is also true for the current study, where just one case was confirmed as deep SSI with a positive bacteriological culture. This also means that the superficial SSI may have an artificially higher value than true values.

The authors have mentioned the mechanical advantages of medial plating over a cranial/dorsal plate. In addition to this, placement of a plate on the medial aspect of the bone offers the advantage of minimal soft tissue dissection when compared with a palmar approach.¹⁸ Thus, a medial plate addresses the disadvantages of both palmar and cranial/dorsal plates and has been reported for PCA.¹⁷ Orthogonal plate application for the canine carpus has been reported for limb-sparing surgery²² and in PCA after highly comminuted articular fractures.²³ Orthogonal plating increases the rigidity of the construct and reduces the strain at the arthrodesis site, increasing the resistance to fatigue failure of the implants.²⁴ Theoretically, by increasing the stability of the construct, cyclical failure of the implant and delayed union should be reduced. With medial plating, multiple metacarpal bones are engaged, stabilizing the carpo-metacarpal joint and adhering to the principle of rigid stabilization of arthrodesis.

Single cranial/dorsal plating engages only one or two metacarpal bones; therefore, this construct is less stiff than orthogonal plating.

Traditionally, rigid external coaptation has been recommended postoperatively in dogs with PCA,^{1,2,5–8,11,17,19} but the application of a cast or splinted dressing has a high incidence of soft tissue complications.²⁹ With the application of biomechanically superior plates, studies have reported good outcomes in dogs with single-plated PCA without the use of postoperative external rigid coaptation.^8,10,30 These studies did, however, have prolonged nonrigid external coaptation. The duration of the application of these bandages after PCA varies between studies. Ramirez and Macias¹⁰ placed a modified Robert-Jones dressing for a mean of 4.8 wk. Chong³⁰ reported a median coaptation duration for the advanced locking plate system group of 51.5 days and 43 days for the hybrid dynamic compression plate group. Bristow⁸ reported that 55.2% of cases with a hybrid dynamic compression plate had external coaptation for a median duration of 42 days. The duration of nonrigid external coaptation in the current study was a mean of 3 days and maximum of 7. Although Tuan et al.¹⁶ applied nonrigid external coaptation for a median duration of 5 days, there was a maximum duration of up to 42 days. Similar to the current study, Tuan et al.¹⁶ used additional fixation alongside the dorsal/cranial plate.

Placement of a dorsal plate with 10–12 degrees of extension means that if a straight plate was used on the medial aspect, it could not extend too distally on the metacarpals owing to the sagittal plane alignment. A longer reconstruction plate was used in some cases, permitting in-plane bending.

The intraoperative complication that occurred in this study is unusual when placing a screw in a standard long bone. The authors believe that the screw head sheared off secondary to the torque used when placing the screw. This is more likely to happen when there is a combination of the following factors: smaller screws with reduced core diameter, the use of self-tapping screws as they have a small cutting flute, and the fact they traverse through multiple cortices. Therefore, it is recommended to tap the drill hole before small-diameter screws are placed.

Fractures of metacarpal bone III have been reported due to the creation of a stress riser when the screw size is too large or when the plate-to-bone ratio is <50%.³¹ Medial plate application, in addition to the aforementioned advantage of stabilizing all metacarpal bones, has been theorized to reduce the risk of a stress riser on metacarpal bone III compared with standard cranial/dorsal plating.^22,23 When applying the medial plate, the authors ensured that the screws engaged metacarpal bone II distal to the proximal 1/3 diaphysis. The concern was that if the screws from the medial plate engaged metacarpal bone III close to where the cranial/dorsal plate ended, there would be a stress riser created from orthogonal bone holes in the mid diaphysis, therefore increasing the risk of metacarpal fracture. In this manuscript, just one dog (1/21, 4.8%) had a fracture of metacarpal bone III; this is in contrast with other publications, where incidence of metacarpal fracture was higher (6/54, 11.11%).³¹ The radiographs of the dog with metacarpal bone III were reviewed. The dorsal plate spanned ∼60% of the third metacarpal bone, the screw size was subjectively adequate, and there were no screws from the medial plate engaging the third metacarpal at the level of the fracture. Therefore, the exact cause of the metacarpal fracture was unknown. After this case, the authors elected to place the most distal screw as a monocortical one, wherever possible.

Postoperative wound dehiscence was observed in one dog (case 20) in this study. Dehiscence of the wound in this case was reported to be due to self-trauma. Potential causes for the self-trauma include suture reaction, implant-associated reaction, skin irritation from clipping, skin tension, infection, etc. The dehiscence of the distal surgical wound was small and healed by second intention, but at a later date, the dog was diagnosed with superficial SSI, which may explain the reason of self-trauma. The suspected SSI resolved after a course of antibiotics. Long-term follow-up is available for this patient and was reported to be excellent with no recurrence of infection.

It has been theorized that the presence of two plates may increase the risk of skin tension, increasing the risk of wound dehiscence and secondary postoperative infection, but this was not the case in the current study. The authors believe that ostectomy of the medial styloid process and the abductor pollicis longus groove, in addition to burring the medial aspect of the base of the metacarpal bone II to make the medial surface flat, helps in closing the surgical incision without tension. For this reason, and to ease contouring, the authors recommend ostectomy of these areas. Alternatively, some surgeons have reported the necessity to amputate digit I.¹⁷ No dog of the current study required digit I amputation.

It has been reported that postoperative antimicrobials in dogs after tibial plateau leveling osteotomy has been protective against developing SSI,³² but the prescription of antibiotics is still controversial in clean elective orthopedic procedures and may contribute to antimicrobial resistance.³³ The application of a second plate will increase the surgical time and the number of implants; therefore, these factors may increase the likelihood of SSI. Even though PCA is considered a clean surgical procedure,³³ some surgeons prescribe postoperative antibiotics owing to the higher incidence of SSI in patients after PCA compared with other elective orthopedic procedures.¹⁶ Surgical site infection was diagnosed in five dogs (23.8%): one was considered deep SSI and had a confirmation of a positive bacterial culture, and the other four were superficial and presumptive. The infection rate in this study was lower than that in the study by Chong et al.,³⁰ similar to Tuan et al.,¹⁶ and subjectively higher than others.^8,10 When SSI is treated, antibiotic therapy may resolve the clinical signs; but in cases of a deep infection, when the antibiotics are stopped, the infection may recur. In these cases, implant removal may be indicated to resolve the infection. Only 1 out of the 21 dogs (4.8%) required implant removal to resolve the infection. Of the 5 dogs diagnosed with SSI, 4 resolved with antibiotic therapy alone. This gave a superficial infection rate of 4 out of 21 dogs (19%). None of the superficial SSI cases were confirmed with bacteriologic testing; therefore, because of the retrospective nature of our study, the infection rate may be artificially higher than the true values and may be why it is higher than the standard orthopedic infection rate of 6.3%.³⁴ Furthermore, none of the superficial SSIs recurred; however, the lack of long-term follow-up of 2/4 of these cases is a limitation of the study.

The radiographic appearance of the arthrodesis healing degree was evaluated retrospectively in all dogs. The healing scales were similar between the intercarpal and carpometacarpal joints, and in both aforementioned joints, the degree of arthrodesis was greater than the radiocarpal joint. It is not clear why the radiocarpal joint may heal slower than the other joints, although it could be theorized that there is a wider joint space, leading to a larger bone gap following the PCA procedure if the surgeon was unable to achieve adequate interfragmentary compression.

Some degree of implant removal was necessary in three (3/21, 14.3%) cases; all implants were removed in two cases (cases 10 and 16), and one case (case 12) had a screw removed from metacarpal bone III. Sawyere et al.⁹ and Tuan et al.¹⁶ reported higher percentages of cases (4/14, 28.6% and 4/12, 33.3%, respectively) requiring implant removal. Ramirez and Macias¹⁰ reported 20% of cases needing implant removal, of which one (1/15, 6.7%) required revision of the arthrodesis at a later stage. Case 10 had persistent mild right forelimb lameness that resolved after implant removal. The exact cause for the lameness was not determined, and although SSI cannot be completely excluded, no bacteria were identified on culture of the implants. Case 16 was diagnosed with immune-mediated polyarthritis 6 mo after surgery. Three months after the diagnosis of immune-mediated polyarthritis, this patient was diagnosed with SSI and required staged implant removal 3 and 4 mo later. Case 12 had a moderate right forelimb lameness due to metacarpal bone III fracture (Figure 2). The lameness improved soon after the most distal screw of the dorsal plate was surgically removed. Of the two cases that had all implants removed in this study, neither required revision of the arthrodesis at a later date. The limitation of this is the lack of long-term follow-up; however, the authors would have expected these patients to have returned if there was fracture of the arthrodesis.

View Figure

• Download as Powerpoint
• Download Figure

Long-term follow-up was available for nine dogs and was assessed by owner CBPI questionnaire. Pain scores were favorable for all cases, apart from one (case 18), which was moderate. In general, functional scores were very good, indicating that dogs with PCA are able to perform normal activities without difficulty. Case 18 was the only patient with fair functional outcome. This patient was diagnosed with elbow dysplasia and secondary degenerative joint disease, which can impact the pain score and function perception in that limb by the owner. The function that scored the highest was “ability to run” (32/90), suggesting that dogs with PCA may not be able to run as well as before surgery. This is not unexpected following PCA, owing to the lack of motion of the carpus. The overall quality of life (assessed by owners) was good to excellent for most of the dogs.

Limitations of this study include its retrospective nature; lack of long-term follow-up in many cases, resulting in incomplete clinical and radiographic follow-up (mentioned through the discussion); the small number of cases reported with no comparator; and subjective outcome measures.

Conclusion

Canine PCA using orthogonal plating is possible without the need for digit I amputation or postoperative rigid or prolonged nonrigid external coaptation and offers good to excellent short-term and long-term outcomes. Despite theoretical concerns that orthogonal plating would increase the risk of SSI and dehiscence as a result of skin tension, in the present study, only one dog (4.8%) required implant removal because of SSI and only one dog (4.8%) had mild dehiscence (which was attributed to self-trauma by the patient). Further studies should aim to compare the various forms of fixation described for canine PCA to facilitate more informed decision-making for surgeons when deciding what technique to use.