Clinical and Radiographic Evaluation of Short- and Long-Term Outcomes of Different Treatments Adopted for Elbow Medial Compartment Disease in Dogs

Medial compartment disease is a common occurrence in dogs affected by elbow dysplasia. Despite many treatments suggested in the literature, only few studies reported comparative outcomes in the short and long term. The aim of this study is to report and compare short- and long-term clinical and radiographic outcomes of dogs treated for medial compartment disease (MCD) by distal dynamic ulnar ostectomy (DUO), bi-oblique dynamic proximal ulnar osteotomy (BODPUO) and conservative management (CM). From 2016 to 2018, all medium to large dogs, aged between 5 and 12 months, affected by uni/bilateral MCD and treated by DUO, BODPUO or CM, were enrolled in this study and followed up for 24 months. Orthopedic and radiographic examinations were performed at T0, T2, T12 and T24 months after treatment. Lameness score, elbow arthralgia, elbow range of motion (ROM), osteoarthritis (OA) score and percentage of ulnar subtrochlear sclerosis (%STS) were evaluated at each time point. According to the treatment performed, dogs were divided into three groups: DUO, BODPUO and CM. Forty-five elbows from twenty-six dogs, treated with DUO (n = 17), BODPUO (n = 17) or CM (n = 11), were prospectively enrolled in the study. The patients enrolled in the CM group were older and showed more severe radiographic signs of OA, compared to those enrolled in the other two groups. Lameness and arthralgia scores (p < 10−4) were significantly decreased in patients that underwent surgical treatment and increased in patients managed conservatively (lameness p < 10−4, arthralgia p = 0.3068), at T12 and T24. OA score (p < 0.0040) and ROM (DUO, CM p < 10−4; BODPUO p = 0.0740) worsened in every study group, but %STS decreased in DUO (p = 0.0108), increased in the CM group (p = 0.0025) and remained unchanged in the BODPUO group (p = 0.2740). This study supports the clinical efficacy of DUO and BODPUO in reducing lameness, arthralgia and progression of %STS. Early diagnosis and surgical attention in patients affected by MCD can improve the short- and long-term outcome and reduce the progression of secondary changes.

Vet. Sci. 2022, 9,70 3 of 19 body condition score (BCS), lameness and arthralgia scores and ROM were recorded at each time point by the same operator.
The degree of lameness and arthralgia was assessed by an expert orthopedic surgeon (A.P.P) using a modified Numerical Rating Scale (NRS) proposed by Vasseur et al. (1995) [35,36] (Table 1). ROM was clinically measured with an orthopedic goniometer [37]. Table 1. NRS used to assess the degree of lameness and arthralgia.
The degree of lameness and arthralgia was assessed by an expert orthopedic surgeon (A.P.P) using a modified Numerical Rating Scale (NRS) proposed by Vasseur et al. (1995) [35,36] (Table 1). ROM was clinically measured with an orthopedic goniometer [37]. Table 1. NRS used to assess the degree of lameness and arthralgia.

Assessment Grade Description
Lameness 0 No evidence of lameness neither at the walk nor at the trot 1 No evidence of lameness at the walk, mild lameness at the trot 2 Mild lameness at the walk, apparent lameness at the trot 3 Apparent lameness at the walk and at the trot 4 No lead of limb during the walk and the trot Arthralgia 0 No pain response 1 Head movement, suspension of breath 2 Subtraction of the limb 3 Vocalizations, aggressiveness OA was assessed with the modified IEWG scoring system using the form by Lang et al. 1998: score 0 = normal (grade 0); score 1 = borderline (grade BL); score 2-4 = mild OA (grade 1); score 5-8 = moderate OA (grade 2); score > 8 severe OA (grade 3) [38]. The same IEWG modified scoring system was used to classify the INC, based on magnitude of radio-ulnar and humero-ulnar steps: mild INC, step < 2 mm; moderate INC, < 4 mm; severe INC, > 4 mm. The subtrochlear sclerosis of the ulna (STS) was measured using a percentage scale (% STS), as previously described [10,[39][40][41] (Figure 1). Figure 1. The %STS was calculated as 100(x/y). The craniocaudal ulnar depth (y) was measured from the most proximocaudal aspect of the radial head to the most caudal margin of the ulnar proximal metaphyseal cortex; the depth of sclerosis (x) was measured from the most proximocaudal aspect of the radial head to the STS caudal border. The figure represents the %STS detected on the same elbow at 0, 2 and 12 months from DUO.

Treatment
The selection of the appropriate treatment for each patient was performed according to the treatment algorithms available in the literature [8,12,27,30].
Four-to six-month-old symptomatic puppies with mild radiographic changes (presence of STS without osteophytes and/or mild INC and MCPD) were treated with DUO surgery. Four-to eight-month-old symptomatic dogs with more severe radiographic changes (presence of STS and moderate INC, MCPD and/or OCD-kissing lesion) were treated with BODPUO surgery. Figure 1. The %STS was calculated as 100(x/y). The craniocaudal ulnar depth (y) was measured from the most proximocaudal aspect of the radial head to the most caudal margin of the ulnar proximal metaphyseal cortex; the depth of sclerosis (x) was measured from the most proximocaudal aspect of the radial head to the STS caudal border. The figure represents the %STS detected on the same elbow at 0, 2 and 12 months from DUO.

Treatment
The selection of the appropriate treatment for each patient was performed according to the treatment algorithms available in the literature [8,12,27,30].
Four-to six-month-old symptomatic puppies with mild radiographic changes (presence of STS without osteophytes and/or mild INC and MCPD) were treated with DUO surgery. Four-to eight-month-old symptomatic dogs with more severe radiographic changes (presence of STS and moderate INC, MCPD and/or OCD-kissing lesion) were treated with BODPUO surgery. Four-to twelve-month-old symptomatic puppies with radiographic signs of severe MCD (severe INC, MCPD and/or OCD-kissing lesion and sign of OA) were treated with CM. CM was performed also in dogs where surgical options were declined by the owners. Conservative management consisted of weight control, a joint-type diet, modulation of onlead exercise and 14 days of oral carprofen (4 mg/kg for 7 days followed by oral carprofen 2 mg/kg for 7 days). Administration of carprofen was repeated as needed.
Patients from DUO and BODPUO groups also underwent a diagnostic elbow arthroscopy. If present, fissure/fracture of the medial coronoid process was arthroscopically removed. A modified Robert Jones bandage was applied for 24 h postoperatively and carprofen (4 mg/kg orally once daily) was administered for 7 days in all dogs. Physical activity was restricted during postoperative phase with lead walks for 2 months. A weight control and joint-type diet were subsequently prescribed.

Statistical Analysis
Degree of lameness and arthralgia, BCS and OA scores were compared between groups using Kruskal-Wallis test followed by Dunn's multiple comparison test, or using Mann-Whitney test, where appropriate, at each time point. Friedman analysis followed by Dunn's test were used to perform a multiple comparison between different time points within each group.
Cardinal data were assessed for normality using D'Agostino-Pearson test. Range of motion and %STS were compared between groups using One-Way ANOVA analysis followed by Holm-Sidak post hoc test or using Student's t-test, where appropriate. A comparison between different time points within each group was performed using Repeated Measures ANOVA followed by Holm-Sidak test.
Statistical analysis was performed in GraphPad Prism, version 8.2.1 (GraphPad Software Inc., San Diego, CA, USA) and p < 0.05 was considered statistically significant.

Animals, Diagnosis and Treatment
Twenty-six dogs (45 elbows) were enrolled in the study. Thirteen breeds were represented: nine Labrador Retriever, three German Shepherd, two Boxer, two White Swiss Shepherd Dog, two Border Collie, one Saint Bernard, one Chow Chow, one Bernese Mountain Dog, one Golden Retriever, one English Bulldog, one Great Dane, one Tchorny Terrier and one American Staffordshire Terrier, for a total of nineteen males and seven females. Nineteen dogs were bilaterally affected and seven unilaterally. Forty-five elbows were enrolled in the study. Twenty-one right and twenty-four left joints were treated. Mean ± SD age at T 0 was 6.7 ± 1.3 months. Mean ± SD BW and BCS were 25.2 ± 7.9 kg and 4.6 ± 0.6, at T 0 , 27.8 ± 8.5 kg and 4.6 ± 0.6 at T 2 , 33.2 ± 10.3 kg and 5.5 ± 0.9 at T 12 , and 36.8 ± 9.9 kg and 6.4 ± 1.3 at T 24 , respectively.
Within the BODPUO group, the radiographic follow-ups showed an excessive proximal ulnar displacement in five elbows [28]. Therefore, the BODPUO group was further divided into the BODPUO-D group, which included five joints with proximal segment dislocated, and the BODPUO-ND group, which included 12 joints with proximal segment not dislocated.
There was no significant difference of BCS between all study groups (p > 0.05), except for the comparison between BODPUO subgroups (p ≤ 0.0123). An improving trend in the BCS was instead appreciated within each group (p ≤ 0.0004) ( Table 4).

Radiographic Examination Findings
The radiographic OA scores significantly increased in all study groups at long-term evaluations (p < 0.05) (Figure 3a). Radiographic signs of OA were more noticeable in the BODPUO-D group than in the BODPUO-ND group at T 12 (T = 5.000; p = 0.0068) and T 24 (T = 4.500; p = 0.0040), because in the BODPUO-D group the score was significantly increased (χ 2 r = 14.47; p < 10 −4 ) at T 12 (p = 0.0373) and T 24 (p = 0.0022) (Figure 3b). From the comparison of DUO, BODPUO-ND and CM groups there was a difference between BODPUO-ND and CM group at T 12 (p = 0.0577) and T 24 (p = 0.0156). At T 12 there was no significant difference between BODPUO and CM groups (p = 0.3289) (Figure 3c) ( Table 5).
In the BODPUO group the %STS remained almost unchanged during follow-up (F = 1.347; p = 0.2740), while in the DUO group it decreased (F = 6.348; p = 0.0108) at T2 (p = 0.0373), T12 (p = 0.0018) and T24 (p = 0.0039). In the CM group it increased (F = 13.20; p = 0.0025) compared with pretreatment values. The %STS was lower in the DUO group com pared to CM group at T12 (p = 0.0110) and T24 (p = 0.0050) and it was lower in the BODPUO group compared with the CM group at T24 (p = 0.0189) (Figure 3d). Comparing the means of %STS in DUO, BODPUO-ND and CM groups, a significant difference was also detected between BODPUO-ND and CM groups at T12 (p = 0.0235) (Figure 3f). The %STS decreased in the BODPUO-ND group (F = 13.33; p = 0.0040) and increased in the BODPUO-D group (F = 0.7874; p = 0.4487). At T24, the BODPUO-ND group showed significantly lower %STS than the BODPUO-D group (p = 0.0481) (Figure 3e).
In the BODPUO-D group, immediate postoperative radiographs were used to meas ure the osteotomy geometry. The means ± SD of the caudo-cranial osteotomy angle and the latero-medial osteotomy angles were 53.2 ± 4.9° and 49.7 ± 1.0°, respectively, while the most caudo-proximal point of osteotomy was situated at 32.7 ± 10.0% of the total ulnar length (Figure 4).     DUO, dynamic ulnar ostectomy; BODPUO, bi-oblique dynamic proximal ulnar osteotomy; CM, conservative management; H, result of Kruskal-Walls statistics; BODPUO-ND, bi-oblique dynamic proximal ulnar osteotomy with proximal segment not dislocated; BODPUO-D, bi-oblique dynamic proximal ulnar osteotomy with proximal segment dislocated; OA, osteoarthrosis; %STS, percentage of subtrochlear sclerosis; T 0 , the day of treatment; T 2 , two months after treatment; T 12 , twelve months after treatment; T 24 , twenty-four months after treatment; X 2 r , result of Friedman statistics.
Vet. Sci. 2022, 9, 70 14 of 19 In the BODPUO group the %STS remained almost unchanged during follow-up (F = 1.347; p = 0.2740), while in the DUO group it decreased (F = 6.348; p = 0.0108) at T 2 (p = 0.0373), T 12 (p = 0.0018) and T 24 (p = 0.0039). In the CM group it increased (F = 13.20; p = 0.0025) compared with pretreatment values. The %STS was lower in the DUO group compared to CM group at T 12 (p = 0.0110) and T 24 (p = 0.0050) and it was lower in the BODPUO group compared with the CM group at T 24 (p = 0.0189) (Figure 3d). Comparing the means of %STS in DUO, BODPUO-ND and CM groups, a significant difference was also detected between BODPUO-ND and CM groups at T 12 (p = 0.0235) (Figure 3f). The %STS decreased in the BODPUO-ND group (F = 13.33; p = 0.0040) and increased in the BODPUO-D group (F = 0.7874; p = 0.4487). At T 24 , the BODPUO-ND group showed significantly lower %STS than the BODPUO-D group (p = 0.0481) (Figure 3e).
In the BODPUO-D group, immediate postoperative radiographs were used to measure the osteotomy geometry. The means ± SD of the caudo-cranial osteotomy angle and the latero-medial osteotomy angles were 53.2 ± 4.9 • and 49.7 ± 1.0 • , respectively, while the most caudo-proximal point of osteotomy was situated at 32.7 ± 10.0% of the total ulnar length (Figure 4).

Discussion
This prospective study reports and compares short-and long-t ographic outcomes in dogs that underwent surgical and conservati age MCD.
Division into study groups was challenging because of the bro presentations and surgical techniques associated with MCD [5,6,16, explains the current paucity of studies that directly compares all tec the decision-making algorithm [8,10,19,27,42]. In the present study, vided according to whether osteotomies were performed and if DU

Discussion
This prospective study reports and compares short-and long-term clinical and radiographic outcomes in dogs that underwent surgical and conservative treatment to manage MCD.
Division into study groups was challenging because of the broad variety of clinical presentations and surgical techniques associated with MCD [5,6,16,[18][19][20][21][22][23][24]. This difficulty explains the current paucity of studies that directly compares all techniques proposed by the decision-making algorithm [8,10,19,27,42]. In the present study, the patients were divided according to whether osteotomies were performed and if DUO or BODPUO were executed.
Ulnar subtrochlear sclerosis is an early sign of elbow dysplasia [30] and has been reported to increase with the progression of the underlying condition [41]. Our results confirmed that the %STS can be useful to assess the progression of the disease, in the short and long term. In addition to that, our results showed that %STS decreased significantly two months after surgery in the DUO group, while it decreased significantly twelve months after surgery in the BODPUO-ND group. This finding supports the hypothesis that DUO and BODPUO may be effective in slowing down the progression of MCD [43]. Our results are consistent with recent clinical studies which showed that the progression of %STS could reduce if the INC is addressed [33,43]. At T 0 , the lameness score in the CM group was lower compared to DUO and BODPUO groups, despite more severe radiographic evidence of OA. However, increased radiographic evidence of OA is not always clinically directly related to the lameness score. It is possible that the increased peri-articular fibrosis associated with the progression of the disease may, to some extent at least in the short term, have increased joint stability and possibly affected the lameness score. In the short term (T 2 ), lameness score in the CM group was significantly lower when compared to the DUO and BODPUO groups. This finding can be explained by the expected postoperative recovery time, following surgical treatment, in the DUO and BODPUO groups. However, the clinical long-term outcome obtained in the DUO and BODPUO groups was significantly superior to the CM group. Lameness and arthralgia scores decreased, in the long term (T 12 -T 24 ), in patients treated by DUO and BODPUO, according to previous clinical studies [30]. The decreased lameness and arthralgia scores observed in our study might be explained by a homogenous re-distribution of the intra-articular loads, following DUO/BODPUO [29].
The ROM decreased and OA score increased in all our study groups. Due to osteophytes and fibrosis interfering with the motion of the joint, moderate inverse correlation between ROM and OA has been previously reported [10,44]. Progression of OA and decreased ROM, in the DUO and BODPUO groups, in spite of an improvement of the clinical outcome, is consistent with previous reviews [10,19,29,45].
A recent study demonstrated that BODPUO does not completely restore the INC and increases the humeroulnar rotational instability [29]. The instability could be responsible for the continuous progression of OA and the absence of a clear improvement in ROM. However, at T 12 and T 24 , the progression of OA and the reduced ROM were significantly lower in the DUO and BODPUO groups compared to the CM.
The ROM of elbows treated by DUO was significantly higher than the mean ROM of BODPUO and CM groups before surgery and at T 2 . This finding was expected, considering that DUO is a surgical procedure recommended in young patients with low cartilage degeneration and with mild clinical symptoms [12,30].
Analyzing elbows treated by BODPUO, we suspected that the excessive proximal ulnar displacement in five elbows had a negative impact on the outcome of the BODPUO group. In fact, at long-term evaluation, the results obtained in the BODPUO-D group were significantly worse than the BODPUO-ND group (excluding lameness and arthralgia score, in which there was no statistical difference).
The mean osteotomy angles and osteotomy position in our study was comparable with what was previously reported by Caron and Fitzpatrick in six elbows with the same complication [28]. In their study, there was no significant difference between the osteotomy angle and position in patients that did and did not develop excessive proximal ulnar displacement. However, the authors supposed that a more acute osteotomy angle or a more proximal osteotomy may lead to excessive motion in some elbows [28]. In a recent study, the authors reported a less severe misalignment than expected if the ulna osteotomy exceeded the recommendations given by Caron and Fitzpatrick [29]. In our study, osteotomies of BODPUO executed too proximally were associated with excessive motion. However, according to recent observations, the excessive motion in our population could also be due to a limited obliquity of the osteotomy [34]. At the long-term follow-up (T 24 ), %STS and OA scores were significantly lower in the BODPUO-ND group than in the BODPUO-D group, while the ROM was significantly higher. Surprisingly, in the BODPUO-D group, despite that the lameness score was not improved at T 12 and T 24 , the arthralgia score was decreased. Considering the low numbers of cases, the subjective assessment of arthralgia, and the theoretical variability of patients' response to conscious examination, this result is of difficult interpretation. Excluding the BODPUO-D group, the result of the CM, DUO and BODPUO-ND groups, in the present study, are consistent with the data reported in literature [28,30].
This study has several limitations. First of all, the older age and the more severe radiographic changes of the patients in the CM group make the statistical comparison of the data with the DUO and BODPUO group questionable. However, our data support the clinical importance of an early diagnosis and the potential benefits associated with dynamic ulnar osteotomy/ectomy, when case selection is appropriate. On the other hand, the present study highlights how conservative management may have a more limited clinical efficacy in older patients with severe radiographic changes associated with MCD.
A second limit is the lack of use of CT osteoabsorptiometry to objectively describe bone density distribution in subchondral bone at the level of the base of the medial coronoid process [46]. In this study CT osteoabsorptiometry was not used, but this limitation allowed to stress the utility of evaluating the %STS as an objective parameter that anyone with X-ray equipment can use [47].
In the present study, the position of the elbows was standardized, in order to allow repeatable measurements of %STS. However, the INC and the presence of osteophytes on the caudal aspect of the radial head may have affected the assessment of the most proximocaudal aspect of the radial head (a reference point to measure the %STS) [40]. The possibility that %STS may vary depending on the dog breed should be taken into consideration [46]. Therefore, it should be kept in mind that %STS is not a parameter usable to compare individual elbows of dogs of different breeds, but it could be an interesting parameter to assess and monitor the progression of the disease after surgical treatment.
Elbow incongruity was classified by the modified IEWG score during radiographic examination at T 0 , in order to define the severity of radiographic signs and establish the appropriate treatment. Nevertheless, the INC was not assessed in the short and long term, because the radiographic exam did not allow an accurate measurement. In fact, in the past decade, the golden standard for incongruity detection was CT, which provides images without overlapping [48,49].
The absence of a CT scan or arthroscopy did not allow characterization/grading of the severity of the disease in the CM group. The group consisted in older patients with severe radiographic evidence of OA or patients in which the owners declined surgical options. Assessing the outcome of conservative management in such a heterogenous population is difficult.
Moreover, it is widely recognized that keeping the patient's body condition score at the low end of the normal range slows the progression of degenerative joint disease and the clinical signs associated with it [50]. All the enrolled dogs increased their BCS throughout the study, thereby the weight gain was a conditioning factor and it could be argued that part of the conservative management in the CM group was not adequately performed and that those patients could have had a better clinical outcome if weight control was implemented.
Several arthroscopic procedures (joint exploration, medial coronoid fragment removal and curettage of the medial compartment) were performed in the DUO and BODPUO groups, as required. The influence of these procedures in the final patients' outcome is unknown.
A kinetic and kinematic evaluation of the patients before and after treatment may have objectively confirmed our pre and postoperative subjective evaluation of the lameness. Finally, long-term follow-up was not available for all the patients.

Conclusions
This study emphasized the beneficial effects of DUO and BODPUO in reducing lameness, arthralgia and extension of %STS in young patients affected by medial compartment disease. These results highlight the possible capacity of early surgical procedures to homogeneously distribute the intra-articular loads and to slow down and reduce the secondary changes. In particular, this was demonstrated for BODPUO [29], but further research should be conducted regarding the humero-ulnar joint kinematics after DUO surgery. Conservative management in older patients with severe radiographic evidence of elbow OA may be associated with a worse short-and long-term outcome.
Funding: This research received no external funding.
Institutional Review Board Statement: Ethical review and approval were waived for this study due to the fact that patients were treated following the current available guidelines for the treatment of medial compartment disease in dogs.