The double-arch modified type-1b external skeletal fixator

Summary Objectives: To describe acute correction of antebrachial angular and rotational limb deformities (ARLD) using a new external skeletal fixator (ESF). Methods: Dogs that were presented with lameness caused by ARLD were treated by radial and ulnar osteotomies and acute realignment. A modified type-1b ESF incorporating double arches (DA-ESF) and a novel connecting configuration facilitated alignment with six degrees of freedom. Bilateral deformities were corrected surgically in the same session. Aseptic preparation of both antebrachii allowed comparison of limb alignment. Radiographic evaluation was performed using centre of rotation of angulation (CORA) methodology. Results: Thirty-five antebrachii (22 dogs) underwent surgery. Postoperative limb function was graded as good (n = 31), fair (n = 2), or poor (n = 2). Persistent medial carpal instability was associated with a suboptimal outcome. Postoperative radiographic images of the frontal and sagittal plane joint angles and elbow-to-carpus translation were compared with values that were reported in previous studies, and were within published reference ranges in most cases. Complications included delayed radial osteotomy union (n = 5), delayed ulnar osteotomy union (n = 2) and implant-associated morbidity (n = 3). Clinical relevance: A practical technique for acute correction of complex antebrachial ARLD is suggested, incorporating a new configuration of ESF. Putative limitations of radio-graphic planning using CORA may be compensated by careful attention to intra-operative visual and palpatory assessment.


Introduction
Asynchronous growth of the radius and ulna is a well-recognized cause of antebrachial angular and rotational limb deformity (ARLD) (1,2). Premature closure of the distal ulnar physis, distal radial physis, or both combined occurs most frequently, al-though premature proximal radial physeal closure or radio-ulnar synostosis can also result in asynchronous growth (3). These processes can occur as a result of trauma, metabolic disturbances or genetic factors (4). The distal ulnar physis is involved in 75% of canine thoracic limb growth disturbances (5). Growth retardation of the ulna typically causes a combination of limb shortening, radial procurvatum, external rotation and carpal valgus (1,2,5). In some dogs, disparity of radial length relative to ulnar length can cause elbow or carpal joint incongruity, or both, which, in turn, can precipitate painful osteoarthritis in addition to the functional lameness caused by abnormal posture (1,2).
In this paper, we report the use of a modified type-1b linear ESF, which incorporates IMEX™ 1/3 arches (double arches [DA]), and features a fully adjustable connecting bar assembly, allowing accurate intra-and postoperative manipulation, with six degrees of freedom. We also report the methods used for preoperative planning, intra-operative assessment of limb alignment, and the postoperative radiographic and functional outcome.

Inclusion criteria
All dogs that were presented to Fitzpatrick Referrals for acute surgical correction of clinically significant antebrachial ARLD between April 2006 and January 2009 were included. Clinical significance was defined as a visible postural lameness (with ARLD being noted by the owners), coupled with signs of pain that were elicited upon manipulation of the elbow or carpal joints, or both. All dogs were assessed and managed operatively by one surgeon (NF). Dogs with significant limb shortening were excluded. Significant limb shortening was defined empirically based on the surgeon's (NF) experience, but was broadly classified as >20% shortening compared to the contralateral limb in dogs with limited growth potential, or premature closure of one or more distal antebrachial growth plates in very young dogs in which >20% reduction in limb length was anticipated. Dogs with incomplete clinical or radiographic records to a minimum of one week after DA-ESF removal were excluded.

Preoperative assessment
In all cases, clinical assessment was performed by the primary surgeon. This consisted of visual gait examination and complete orthopaedic examination. Function and limb alignment were separately graded as fair or poor, since all patients were lame and all were affected by valgus and rotation of the distal antebrachium and carpus.
Preoperative radiographs were obtained with the dogs under deep sedation using either medetomidine a and butorphanol b or routine general anaesthesia. Direct digital radiography c was employed with commercially available software d . Radiographic images included medio-lateral and craniocaudal projections with positioning referenced from the transverse axis of the elbow and carpal joints, and collimated to include the entire antebrachium. Measurements were performed following image calibration using the measurement tools intrinsic to the image software. Frontal plane alignment (FPA) and sagittal plane alignment (SPA) measures were calculated as previously described by Fox and others (7). Centre of rotation of angulation measurements were recorded using a previously described technique for dogs with complex deformities (21).

Pre-surgical preparation
Under routine general anaesthesia, dogs were positioned in ventral recumbency, with the head and neck suspended dorsally away from the operative site using a fabric sling. Anaesthetic monitoring was performed using an oesophageal stethoscope, capnography, electrocardiography, pulse oximetry, and indirect (oscillometric) blood pressure measurement using a cuff positioned over the dorsal metatarsal artery. Both thoracic limbs distal to the midhumerus were aseptically prepared and draped into the surgical field. Bilateral deformity cases underwent single-session bilateral procedures. Cefuroxime e (22mg/kg) and carprofen f (2mg/kg) were administered intravenously 20 minutes prior to the start of surgery.
The surgical sequence was: (1) ulnar osteotomy, (2) radial ostoeotomy, (3) ESF transfixation of the proximal and distal radius, and (4) application of the connecting apparatus. All osteotomies were created using a pneumatic microsagittal saw g and a 5.8 mm saw blade h with saline coolant. Intra-operative fluoroscopy was not used. The surgical wounds were closed routinely, in three layers.

Ulnar osteotomy
An assistant elevated and abducted the antebrachium to expose the caudal antebrachium, allowing a limited caudal approach to be made to the proximal ulnar diaphysis (22). A dynamic proximal ulnar osteotomy was made in all cases, as previously described (22,23). Briefly, the osteotomy was directed obliquely from caudoproximal to craniodistal and proximolateral to distomedial (20-40 degrees to the long axis in both trajectories), at a level approximately one-third of the distance from the humero-radial joint to the radio-carpal joint.

Radial osteotomy
A transverse radial osteotomy was performed at the point of maximum deformity as determined intra-operatively, by visual assessment, by trans-cutaneous palpation of the radius, and as identified as the distal CORA on the preoperative radiographs. Position of the osteotomy was typically at the level of two-thirds of the radial length from proximal to distal. frame was comprised of two one-third ring arches (50, 66 or 84 mm internal diameter depending on limb size), one proximal and one distal to the radial osteotomy. The arches were orientated transversely, with the central hole positioned directly dorsal to the radius. Each radial bone segment was secured with three to five negative threaded Ellis pins j attached to an arch component. Transfixation pin diameter was determined by metaphyseal diameter, with a target size of 20-25% bone diameter. Pin diameters of 1.1 mm, 1.6 mm, 2.0 mm, 2.4 mm, 2.7 mm, 3.0 mm or 3.2 mm were used. Pins were placed without pre-drilling, but with the use of saline irrigation throughout.

Modified type-1b linear external skeletal fixator with incorporated double arches
The first pin was placed from craniomedial to caudolateral in the distal radial fragment, at a level approximately 7 mm from the radial ostoetomy. The distal onethird arch was attached to this pin using a half-pin clamp, and the following two to four Ellis pins were placed convergently through half-pin clamps or SK™ clamps pre-mounted on the distal arch. The second pin placed was positioned as far distally as possible, but not within 5 mm of the antebrachiocarpal joint, so that the transfixing elements would span the fragment. One or more of the following connecting clamp variations was used in order to prevent all of the convergent pins from a single arch from transfixing the radius at the same proximo-distal level (ǠFig. 1): 1. Use of a half-pin fixation bolt on either side of the arch (ǠFig. 1A). 2. Application of one or more washers subjacent to a half-pin fixation bolt to elevate the bolt from the arch (ǠFig. 1B). 3. Application of a 6 mm threaded rod as an extension post on the arch (with a 6 mm threaded SK™ clamp for pin attachment) (ǠFig. 1C). 4. Application of a half-pin fixation bolt to a one-or two-hole post elevated from the arch (ǠFig. 1D).
Pin sequence for the proximal arch was performed similarly to that described above for the distal arch, with the exception that since the proximal radial segment was longer, the most distal pin was further from the osteotomy and the most proximal pin was placed craniolaterally at sufficient distance from the humero-radial joint to assure free range of movement of the elbow joint without impingement of the frame on the cranial aspect of the humerus in flexion. Transfixation pins were inserted obliquely from the cranio-lateral and cranio-medial aspects of the bone rather than true lateral or medial, such that they did not traverse the ovoid proximal radial cross-section at the narrowest transverse dimension. Care was taken to limit pin placement to previously defined safe corridors with particular emphasis on avoidance of the extensor carpi radialis muscle craniomedially (24). Pins were applied proximal to the distal radial physis and distal to the proximal radial physis in skeletally immature dogs.

Connecting complex assembly 'and limb realignment
The connecting elements consisted of one of the two following systems, both of which allowed fragment manipulation with six degrees of freedom (ǠFig. 2): 1. A one-or two-hole post (depending on patient size) was mounted off the medial aspect of the proximal arch and the lateral aspect of the distal arch with a one-hole post attached perpendicularly, in which a 6 mm half-pin fixation bolt was placed. A 4.8 mm Steinmann pin was positioned as a connecting rod through the fixation bolts (ǠFig. 2A).

A Universal SK™ Hybrid Adapter k was
attached to each arch element and a carbon fibre connection rod positioned through the split-ball-bearing elements of the devices (ǠFig. 2B). The adapter was attached to the medial aspect of the proximal arch and the lateral aspect of the distal arch; this was done because all cases required medial rotation and angulation of the distal segment. The Hybrid Adapters allowed connection rod placement with 65° of adjustability about its central axis. Acute antebrachial realignment was performed as referenced from the frontal plane  Photographic images of the fully adjustable connecting apparatus used to connect the proximal and distal arches. of the humero-radial and radio-carpal joints (ǠFig. 3). Realignment was facilitated by the positioning of the patient in ventral recumbency. This allowed palpation of the transverse plane of the humeral epicondyles and the radial and ulnar styloid processes whilst the elbow and carpal joints were manipulated through their normal ranges of movement. Rotational and frontal plane deformities were sequentially corrected initially by using the distal frame arch component as a handle to apply torsional and latero-medial force. Where intrinsic musculo-tendinous constraint permitted, partial correction of pro-curvatum was performed. Complete correction of pro-curvatum was not possible in all dogs due to excessive tension within the flexor musculotendinous units. Careful judgement was required to achieve sagittal colinearity of the elbow and carpal central axes as much as was possible whilst maintaining adequate bone contact for healing at the osteotomy. Once satisfactory alignment was achieved and maintained by the primary surgeon, proximal and distal frame components were secured by a surgical assistant via sequential tightening of the proximal and distal pre-placed multi-directional connecting element clamps. After appropriate realignment was established, all frames were reinforced by placement of one or two-hole posts on the opposite ends of the arches relative to the multidirectional de-vice, and these posts were connected using half-pin fixation bolts and a 4.8 mm Steinmann pins (ǠFig. 3).

Postoperative care
The DA-ESF was covered with self-adhesive elastic tape to prevent self-trauma on frame components or cut pin ends. Postoperative analgesia consisted of methadone l (0.3 mg/ kg intramuscular every 4 hours) or buprenorphine m (0.02 mg/kg intramuscular every 8 hours) for one to three days. Carprofen was administered postoperatively at 2 mg/kg orally twice daily for one week, reducing to 2 mg/kg orally once daily for two weeks thereafter. The pin-skin interface was allowed to form a cutaneous scab, and no pin-skin interface cleaning was adopted. Cage rest and lead-only walking of progressively increasing duration (10 minutes four times daily to 30 minutes three times daily) were recommended until signs of radiographic union were documented and DA-ESF removal was accomplished.

Clinical reassessment
Clinical reassessment was performed by the primary surgeon every two weeks postoperatively until DA-ESF removal, then one week after frame removal and thereafter as indicated until 24 weeks postoperatively or until lameness resolved in all cases. Clinical examination was subsequently repeated 12-18 months postoperatively where patients were available. Recorded findings included subjective evaluations of lameness and limb function (including assessment of alignment), pain or laxity on joint manipulation, and the presence and nature of any complications.
Postoperative functional outcome was graded using a simple four-point ordinal scale based on clinical notes regarding the combined surgeon and owner assessments of lameness and video recordings (where available): 1. Excellent = functionally normal; 2. Good = mild lameness only after extensive exercise; 3. Fair = persistent mild lameness but consistently weight-bearing; 4. Poor = persistent moderate or severe lameness with or without occasional non-weight-bearing lameness.
Limb alignment following DA-ESF removal was graded using a simple threepoint ordinal scale: 1. Excellent = similar in appearance to the unaffected contralateral limb where pres- ent or no subjectively detectable deformity from anticipated breed appearance; 2. Fair = minor visual difference from the unaffected contralateral limb where present or noticeable deviation from anticipated breed appearance; 3. Poor = marked visual disfiguration compared with unaffected contralateral limb or severe deviation from anticipated breed appearance.

Radiographic reassessment
Radiographic examination was performed using the same positioning, projections and imaging analysis techniques as defined in the preoperative radiographic protocol. Radiographs were obtained immediately postoperatively and every two to three weeks postoperatively thereafter until DA-ESF removal. Where available, radiography was repeated at the 12-18 months reassessment. Translation at the level of the radial osteotomy was measured from the postoperative craniocaudal radiographs using a previously described technique as the distance between two points (25): a) The intersect of the distal radial anatomic axis line and the osteotomized surface of the distal radial segment. b) The intersect of the proximal radial anatomic axis line with a perpendicular line drawn at the same proximo-distal level as point 'a' .
The measured distance was expressed as a percentage of the bone diameter of the distal segment at the osteotomy. Elbow-tocarpus translation was measured from immediate postoperative radiographs using a previously described methodology (21).
Delayed union of the osteotomy was categorized as time to radiographic osseous union exceeding 10 weeks.

Data analysis
Statistical comparisons for radiographic data (bone segment translation, frontal plane alignment [FPA], sagittal plane alignment [SPA], distal centre of rotation of angulation [CORA] measurements in the frontal plane, and time to radiographic bony union) were performed using a statistical software package n . Radiographic measurments of FPA and distal CORA taken preoperatively and immediately postoperatively were compared using paired T-tests. A multiple regression model was performed to analyze the potential associations of age and body weight with time to bone union. Confidence interval was set at 95%.

Preoperative clinical assessment
All dogs were presented with thoracic limb lameness and angular limb deformity identified as carpal or distal antebrachial valgus with external rotation (supination) of the carpus and manus in 33/35 antebrachii, and as external rotation of the distal radius, carpus and manus alone in two out of 35 antebrachii. All dogs were affected by radial procurvatum and relative flexor tendon tautness and carpal valgus. There were variable degrees of medial carpal laxity and pain induced by forced carpal extension or application of varus/valgus stress to the manus. Clinical examination also revealed signs of pain upon manipulation of the ipsilateral elbow in all cases, especially in extension or in pronation/supination. According to the pre-determined subjective scoring system, limb function was graded as 'fair' (n = 21) or 'poor' (n = 14). Limb alignment was graded as 'fair' (n = 10) or 'poor' (n = 25). In all cases, although ARLD had been noted prior to presentation, manifestations of pain as perceived by the owner (including exercise intolerance or spontaneous vocalisation) were the primary reason for presentation.

Preoperative radiographic assessment
Frontal plane alignment, SPA and CORA measurements recorded preoperatively are documented in ǠSupplementary Table 1 (Available online at www.vcot-online. com). Radiographic examination of skeletally immature dogs revealed premature closure of the distal ulnar growth plate, the distal radial growth plate, or both. Overt elbow incongruity (identified on orthogonal radiographs) manifested as humeroulnar subluxation was present in 16/35 elbows. Rotational deformity (defined as oblique positioning of the carpus when the elbow was appropriately positioned in the frontal plane on cranio-caudal radiographs) was present in all antebrachii.

Surgical findings
Three transfixation pins were applied to the distal segment in 28 antebrachii and four pins were applied in seven antebrachii. For the proximal segment, three fixation pins were applied in eight antebrachii, four in 17 antebrachii, and five in 10 antebrachii. Pin numbers applied per segment were determined by two factors: maximal pin diameter that was accommodated by the radial cross-sectional area (narrower pins generally requiring increased number to provide adequate structural rigidity; particularly a feature of toy-breeds) and the proxi-mal to distal separation distance between pins as permitted by available bone length. Mean surgical time was 40 minutes (range 35 to 50 min) for unilateral procedures and 106 minutes (range 80 to 140 min) for bilateral procedures.

Postoperative clinical assessment
Presence of the DA-ESF did not preclude nearly-normal ambulation, even in bilaterally operated cases. Dogs were able to walk from the day after surgery, and many owners commented that exercise tolerance was improved immediately after discharge (despite imposed restrictions). When bilateral frames did contact one another during normal gait, this was noted as a mild intermittent problem only. In cases affected by the complications of delayed union or major pin tract inflammation, only one dog suffered transient lameness, which improved after removal of a loose pin. One week after DA-ESF removal, signs of discomfort on elbow and carpal manipulation were either absent or slight in all cases. Limb function after DA-ESF removal was graded as 'good' (n = 31), 'fair' (n = 2) or 'poor' (n = 2). These findings were maintained at subsequent reassessments in all cases, with a median follow-up duration of 14.5 months (range 12-18 months). Persistent medial carpal laxity and resultant valgus posture was apparent in both dogs having a 'poor' functional outcome. Proximal and distal radial articular surfaces were satisfactorily aligned for both of these cases and postural valgus was attributed to medial collateral ligament laxity.
Limb alignment (normal standing posture) was graded as 'excellent' in 33/35 antebrachii (ǠFig. 5) one week after DA-ESF removal and 'poor' in the two of the 35 antebrachii with persistent medial carpal laxity (1 Bassett Hound, 1 Shih Tzu). Alignment was maintained until the final followup examination in all cases.
There was no significant difference between radiographic measurments of limb alignment immediately after frame removal by comparison with alignment measurments immediately postoperatively (p <0.001). Mean time to radiographic union and DA-ESF removal was 9 ± 3.8 weeks (range 5 to 18 weeks). Delayed union of the radial osteotomy occurred in five limbs, with concurrent delayed union of the ulnar osteotomy in two out of five. In all cases, this required prolonged maintenance of the DA-ESF. Although case numbers were too small to allow statistical comparison, there did not appear to be a greater degree of lameness in dogs affected by delayed radial or ulnar union when the DA-ESF was in place. There was no significant effect of age (p = 0.3) or weight (p = 0.4) on time to radiographic bone union. Four out of five dogs with delayed union had undergone bilateral surgical correction, but the group size was too small to analyze the effect of single-stage bilateral surgery on ulnar osteotomy healing with 80% power. Radiographically detectable elbow incongruity was not seen in any case at the time of DA-ESF removal.
Nine dogs (13 antebrachii) were available for 12-18 months postoperative radiographic reassessment. As before, no evidence of elbow incongruity or progression of periarticular osteophytosis (assessed subjectively by the lead author) was seen at this time.

Complications
No intra-operative complications were encountered. Postoperative complications were recorded in eight of 35 antebrachii and included delayed union of the radial osteotomy (n = 5) with concomitant delayed union of ulnar osteotomy (n = 2) and major pin tract inflammation, defined as visible discharge with or without transfixation pin loosening (n = 3; 1 pin was palpably loose). Dogs affected by major pin tract inflammation were not the same as those affected by delayed union. All pin tract complications occurred within the last two weeks of frame application. Major pin tract inflammation necessitated removal of the affected pin (not replaced) and antibiotic therapy in one case, while two cases were managed with antibiotic therapy only. Major pin tract inflammation was exclusively associated with excessive cranial orientation of the most proximomedial pin of the proximal segment adjacent to the muscle belly of the extensor carpi radialis. All complications resolved uneventfully following initial therapy and did not have any effect on eventual clinical outcome. Minor pin tract inflammation was occasionally observed, but in no case did this require systemic medication, implant removal or DA-ESF modification.

Discussion
Antebrachial angular and rotational limb deformity is a particularly challenging problem for the veterinary orthopaedic surgeon in terms of the decision-making processes, preoperative planning, and techniques employed for surgical intervention. The application of the DA-ESF was considered straightforward compared with previously described techniques for addressing complex ARLD, and surgical duration was typically less than one hour per limb. Short-term functional outcome and complication rate was at least equivalent to existing techniques, and ESF morbidity was limited, with no reported failure of fixation in any case.
Comparison of the DA-ESF surgical technique to pre-existing surgical options highlights several practical advantages. Surgical fixation options after acute antebrachial realignment constitute internal plate and screw fixation (6) or external skeletal fixation (15)(16)(17)(18)(19)(20). Technique is influenced by the osteotomy/ostectomy shape; the quality and quantity of distal bone stock; clinical factors, including patient temperament and owner circumstances; and personal preference of the surgeon (1,3,(15)(16)(17)(18)(19)(20). Plate and screw application may be especially difficult where derotation is necessary, particularly with respect to conforming the plate to elliptical bone segments in different planes, even when locking screw technology is employed. External skeletal fixation can be readily applied to stabilize any configuration of osteotomy/ostectomy regardless of the degree of de-rotation. Linear ESF configurations are defined by the type and plane of the transfixation pins, and the number of connecting bars (26). Bilateral constructs (types 2 and 3) have a mechanical advantage in terms of stiffness, but suffer a disadvantage in the ability to respect the antebrachial safe corridors. These safe corridors have been described previously, with the best areas for radial transfixation situated proximo-laterally and disto-medially (24). Thus, it is not possible for any bilateral ESF configuration to avoid hazardous corridors, and traditional unilateral (Type 1a and 1b) ESF configurations can only exclusively occupy safe corridors if they have a spiral (free-form) shape. The DA-ESF has a modified type 1b configuration, whereby pins can be positioned to avoid hazardous corridors and additionally avoid transfixation of the transverse crosssectional area of the radius which may be narrow by comparison with pin diameter. It is conceded that not all of the converging transfixation pins from the proximal arch avoided the proximo-medial antebrachial muscle group, and in three cases early pin loosening was noted in this location. Thus, we recommend positioning the most cranio-medial pin from the proximal arch as laterally as possible. This manoeuvre is facilitated by the fact that each 1/3 arch has a total of 11 available holes.
Over recent years, circular and hybrid ESF have gained popularity for the treatment of antebrachial ARLD (17)(18)(19)(20). Thin tensioned transfixation wires allow capture of very small metaphyseal fragments without sacrificing construct stability (17,18). In addition, circular ESF can be used for progressive correction of deformities and limb lengthening (27). These techniques require meticulous planning, extensive experience, frequent postoperative monitoring and, in some cases, multiple surgical procedures (27). None of the dogs in this series were affected by clinically significant limb shortening, so neither progressive deformity correction nor limb lengthening was required. By comparison, the DA-ESF can be quickly and easily constructed intra-operatively and has a significant advantage in the available range of fragment manipulation using the pin-arch constructs as rigid anchors. In our opinion, the IMEX™ Universal SK™ Hybrid Adapter was technically easier to use with regard to acute realignment than the system utilizing one-and two-hole posts and half-pin fixation bolts. There was a lower tendency for the components to 'bind' during three-dimensional manipulation, and the tightening mechanism was simpler. Regardless of which connecting system is employed, we consider it mandatory to employ a secondary reinforcing element (using one-or two-hole posts and half-pin fixation bolts). The constructs employed for this case series facilitated progression to bony union of all osteotomies but biomechanical evaluation has not been performed at this time.
Several recent publications have stressed the importance of preplanning for antebrachial ARLD correction using the CORA methodology, with less emphasis on the use of intra-operative assessment of alignment using visual and palpatory evaluation (17,21,28). Although we collected the data necessary for preoperative assessment using the CORA methodology, and applied it to establish the location of the CORA, the decision to preferentially use intra-operative alignment assessment was premeditated. Importantly, all dogs in this series were affected by marked rotational antebrachial deformity in addition to angular deformity. It is intuitive that single plane estimation of SPA and FPA is markedly affected by the presence of rotational deformity. For example, if an antebrachium with 90° of external rotation of the carpus relative to the elbow is de-rotated by application of an internal rotation of 90°, preoperative measures of angulation in the frontal plane become applicable in the sagittal plane (i.e. a preoperative varus deformity becomes a recurvatum deformity). Also, estimation of radial torsion using radiographic landmarks on the proximal and distal radial articular surfaces is limited because it does not account for either pathologic lateral subluxation of the radial head or normal rotation of the antebrachium around the anatomic axis of the radius. Either feature has the potential to increase the degree of rotation of the car-pus relative to the elbow joint in vivo. We performed a recent ex vivo pilot study to test the validity of the CORA methodology using solid bone models derived from CT scans compared with conventional radiographic measurements (29). This study revealed that radiographic measurements did not accurately reflect actual FPA or CORA measurements in the majority of limbs affected by complex rotational deformities. In contrast, acute correction of rotational deformity in dogs has been described based on intraoperative subjective assessment in both the antebrachium and tibia and there is evidence that clinical evaluation of rotational axial alignment, including use of a torsiometer, may be as accurate as computed tomography measurement for the human tibia, where torsion is normally present (4,21,27,(30)(31)(32). Palpatory or visual evaluations of human long bone torsional parameters are described for both clinical and research applications and are frequently applied for pre-, intra-and postoperative evaluation of patients where surgical correction of rotational deformity is undertaken, frequently in preference to computed tomography or advanced imaging (Personal Communication: Dr Matthew J. Barry, MS FRCS(Orth), Royal London Hospital, London, UK, February 2010) (33,34). In this case series, the objective of realignment of the transverse functional mechanical axis of the humero-radial and radio-carpal joints was achieved in 25/35 limbs, and this rate of complete ARLD correction compares favourably with studies in which radiographic planning has been performed using the CORA methodology (7,21) .
In this case series, we deviated from the conventional patient preparation technique in order to optimize both intra-operative assessment of alignment (with the contralateral limb included in the sterile field) and technical ease for osteotomy creation and DA-ESF application (with the patient in ventral recumbency). Whilst this technique was employed safely and effectively in all cases, anaesthetic monitoring required particular care due to the restricted access to the patient's head. A similar situation is commonly encountered in ophthalmological surgery, and the anaesthetic challenges and recommended pre-cautions for these patients are well described elsewhere (35).
The incidence of preoperative elbow incongruity in this report was similar to that previously reported and, in agreement with Theyse's findings, did not appear to be a significant prognostic factor with regard to short-term clinical outcome (4). A previous study has also demonstrated that proximal (rather than distal) ulnar osteotomy was required to positively influence elbow congruity in an ex vivo model (36). The bi-oblique dynamic proximal ulnar osteotomy described has been reported previously as part of the surgical intervention for humero-ulnar joint pathology in dogs with elbow incongruity (22,23). This technique has the potential to positively influence humero-ulnar incongruence in multiple planes, although this biomechanical effect has yet to be proven definitively. In this series, neither osteo-tomy nonunion nor persisting radiographic evidence of elbow incongruence was observed as a complication after the bi-oblique dynamic proximal ulnar osteotomy.
To simultaneously address humeroulnar incongruity and distal antebrachial ARLD, the ulnar osteotomy was performed proximal to the radial osteotomy. However, whilst we recognize that soft tissue constraints, including the flexor tendons and the interosseous ligament, may restrict segmental motion, the rigid anchor provided by the distal pin-arch construct allowed appropriate distal segment manipulation, and juxtaposition of the ulnar segments was adequate for bony healing in all cases. This may be partly attributable to the fact that de-rotation around the axis of the antebrachium, rather than acute mediolateral bending, was the primary mechanism for correction of the deformities in this case series. It is noteworthy that the two cases of delayed ulnar osteotomy healing were both associated with simultaneous correction of particularly large angular and rotational deformities.
We performed a single radial osteotomy, even in 20 cases where two CORA were identified, since the distal CORA in all cases was of significantly greater magnitude and since our pilot study revealed that the proximal CORA was artefactually exaggerated using a conventional radiographic tech-nique (29). Execution of a single osteotomy, whilst facilitating ease of application, resulted in inevitable elbow joint-to-carpus translation in many cases, with five of 35 limbs affected by more than 20 mm of translation and a maximum of 31 mm in one case. Joint translational deformity may be important in terms of joint loading and overall limb kinematics, but this was not manifested as overt disability of function or evidence of pain in the short term within this case series (34). Dismukes et al. reported that postoperative elbow-to-carpus translation of 26 mm in one dog was not overtly associated with suboptimal functional outcome (21). However, relative joint malalignment and resultant changes in load-bearing force orientation may contribute to osteoarthritis and long-term monitoring in this regard would be a useful adjunct to this study.
We have described a practical and versatile technique for the surgical management of antebrachial ARLD, and highlighted potential benefits of the preferential use of intra-operative assessment of limb alignment. The new DA-ESF configuration allows realignment and osteotomy stabilization to be performed in a single surgical step. Short-term functional outcome was good and was maintained in all dogs followed long-term except for two dogs, with ongoing lameness thought to be a consequence of carpal medial collateral ligament laxity. Although this particular problem has not been documented previously within the veterinary literature, it has been reported anecdotally, regardless of surgical technique. Early recognition of this problem might prompt an attempt to 'over-correct' the valgus malalignment in an attempt to unload the medial collateral ligament. This adaptation would be possible with adjustable ESF configurations, and confers an additional benefit to the DA-ESF. We feel that clinical application of DA-ESF for surgical management of antebrachial ARLD can be recommended.