Comparison of radiographic osteoarthritis scores in dogs less than 24 months or greater than 24 months following tibial plateau leveling osteotomy

Abstract

Radiographic osteoarthritis scores were determined in 60 dogs up to 3 years following tibial plateau leveling osteotomy. Radiographs taken immediately following surgery and at long-term follow-up 1 to 3 years later were evaluated by 2 certified radiologists using a modified 32-point osteoarthritis scale. Changes in osteoarthritis scores were evaluated by paired t-tests and regression analysis. Sub-groups of dogs were formed to evaluate if osteoarthritis scores changed differently for follow-up periods of < 24 months compared with those > 24 months. There was a significant increase in osteoarthritis score from post-operative to follow-up evaluations for all dogs studied. Regression analysis of the relationship of months after surgery to change in osteoarthritis score was not significant. These results indicate that osteoarthritis did progress following tibial plateau leveling osteotomy, but not in a linear fashion over time.

Introduction

The canine cranial cruciate ligament (CCL) functions to oppose stifle hyperextension, internal rotation of the joint, and cranial tibial translation with respect to the femur (1,2). The CCL is commonly injured, as a result of excessive joint forces or degenerative weakening. Many surgical procedures have been described to overcome the resultant craniocaudal instability, and surgical technique tends to depend on the surgeons’ training and experience (3). Tibial plateau leveling osteotomy (TPLO) was designed to eliminate active weight-bearing cranial tibial thrust (2). The goals of TPLO include return to full range of motion, return to normal limb musculature, elimination of joint inflammation, cessation of progression of osteoarthritis, and full return to function (2).

Osteoarthritis (OA) associated with CCL rupture is thought to be secondary to trauma or degeneration leading to joint instability, and has been shown to progress despite treatment modality (4–12). Radiography has been the diagnostic tool most frequently used in veterinary medicine to evaluate OA (13). Several grading systems have been developed to quantify OA changes (8,10,14,15). Radiographic changes include joint effusion, osteophytosis, enthesiophytosis, intra-articular mineralization, subchondral sclerosis, subchondral cyst formation, and narrowing of joint space (15). Although the progression of OA following TPLO has been demonstrated, the relationship between time and OA changes with long term follow-up has not been established.

The purpose of this study was to evaluate whether radiographic OA changes in dogs up to 3 y after TPLO surgery are progressive and time-dependent. We hypothesized that osteoarthritis will progress over time in a linear fashion.

Materials and methods

Patient selection

Medical records of dogs presented for TPLO at Mission MedVet between May 2005 and May 2008 were evaluated. Cases were excluded if the medical record was incomplete, if preoperative radiographs were not available, if preoperative or follow-up radiographs were deemed of inadequate quality by either of 2 board certified radiologists, if the patient was not available for follow-up examination, or if there was evidence of concurrent orthopedic disease based on orthopedic examination at initial evaluation or at the long-term follow-up visit.

Radiographic analysis

Mediolateral and craniocaudal radiographs of the affected stifle were obtained under general anesthesia immediately after TPLO surgery. For the mediolateral projection, the tibia was placed parallel to the radiographic film and the femoral and tibial condyles were superimposed. The craniocaudal view was centered on the patella which was centrally positioned in the trochlea. The same views were obtained in awake patients at 1 long-term follow-up examination, between 1 and 3 y after surgery. Radiographs were reviewed by 2 board certified radiologists and OA score was determined based on a modified 32 point OA scale (10) (Table 1). We excluded evaluation of subchondral sclerosis based on results of the study by Innes et al (5). A total of 24 locations were evaluated with a score of 0–3 assigned at each location. A score of 0 indicated no degenerative changes were apparent, 1 indicated mild changes, 2 indicated moderate changes, and 3 indicated severe changes. All values were summated, with a possible total score of 0–72. The OA score determined by each radiologist was averaged for each patient and used for statistical analysis.

Table 1.

Twenty-four-point scale used to grade factors associated with osteoarthritis

Factor	Score
Periarticular osteophytes, lateral femoral condyle
Periarticular osteophytes, medial femoral condyle
Osteophytes, femoral intercondylar region
Lateral collateral ligament enthesiopathy
Medial collateral ligament enthesiopathy
Lateral soft tissue thickening
Medial soft tissue thickening
Periarticular osteophytes, proximolateral tibia
Periarticular osteophytes, proximomedial tibia
Osteophytes, central tibial plateau
Meniscal mineralization
Intra-articular mineralized osseous fragments
Intercondylar avulsion fracture fragments
Apical patellar osteophytes
Basilar patellar osteophytes
Stifle joint effusion or capsular thickening
Periarticular osteophytes, femoral trochlear groove
Distal femoral condylar remodeling
Periarticular osteophytes, fabellae, lateral and medial gastrocnemius and popliteal sesamoids
Cranial apical patellar enthesiopathy
Periarticular osteophytes, cranioproximal tibia
Periarticular osteophytes, caudoproximal tibia
Tibial condylar remodeling
Femoral subtrochlear lysis
Total score

Each factor was assigned a score of 0–3 on pre-operative and follow-up radiographs. 0 = no degenerative changes were apparent, 1 = mild changes, 2 = moderate changes, 3 = severe changes.

Statistical analysis

Descriptive statistics were calculated for interval variables recorded in this study, including radiographic scores for OA at postoperative and follow-up periods, age, and weight. The frequency distribution of breed and gender were determined to characterize dogs included in the study. For dogs having surgery on both stifles, each stifle was evaluated individually. Changes in OA scores over time between postoperative and follow-up evaluations were evaluated by paired t-tests and regression analysis. Sub-groups of dogs were formed to evaluate if OA scores changed differently for follow-up periods < 24 mo (group 1) compared to those > 24 mo (group 2). These sub-groups were based on a 24-month period to better equilibrate the number of individuals in each group. Regression analyses included consideration of follow-up time after surgery in addition to other variables, such as weight and gender, for evaluating the relationship between these factors and changes in OA scores. A forward, step-wise variable selection procedure was used as the method for selecting independent variables to produce the best fitting models. For all dogs in the study, as well as the subgroups based on the 24-month follow-up dividing point, the number of months after surgery when the follow-up OA scores were obtained was also placed into the regression equations to evaluate the effect of time on changes in OA scores. Consistency between the 2 radiologists was evaluated with correlation coefficients and t-tests. Unless stated otherwise, a P-value of 0.05 was used to determine significance. Values are reported as mean +/− standard deviation (SD).

Results

Sixty of 435 dogs presented for TPLO surgery between May 2005 and May 2008 fit the inclusion criteria for this study. The mean age at time of follow-up was 7.83 y +/− 2.78 y (SD). Mean weight was 33.66 kg +/− 15.88 kg. Most common breeds represented included Labrador retrievers (n = 16), mixed breed (n = 8), west highland white terriers (n = 4), and German shepherds (n = 3). There were 30 spayed females, 27 neutered males, and 3 intact males. Of the 60 dogs in the study, 29 had only 1 TPLO on 1 stifle, and 31 had a TPLO performed on both stifles at the time of long-term follow-up. Ten dogs had bilateral TPLO surgery during 1 anesthetic episode. For all dogs with 1 surgery date, mean follow-up time was 30.88 mo +/− 9.59 mo. For dogs that had surgery on both stifles at 2 separate surgery dates, mean follow-up time was 25.08 mo +/− 11.85 mo.

For the first stifle operated, mean postoperative OA score was 12.87 +/− 6.85. At follow-up examination, mean OA score was 22.28 +/− 9.28. The mean change in OA score was 9.41 +/− 7.78. For the 21 dogs that had surgery on the second stifle at a second date, postoperative mean OA score was 15.77 +/− 8.37. At follow-up examination, mean OA score was 23.48 +/− 10.20. The mean change in OA score was 7.71 +/− 8.28. The difference in mean change between dogs having surgery on 1 stifle and dogs having surgery on both stifles was not significant. For all dogs in this study, there was a significant increase in OA score from postoperative to follow-up evaluations (P-value < 0.001).

Regression analysis of the relationship of months after surgery to change in OA scores was not significant (P = 0.063). There was not a significant linear relationship between time after surgery and OA score change. When divided into follow-up times of < 24 mo (n = 33) and > 24 mo (n = 57), regression analysis showed there was not a significant difference between these 2 groups (P = 0.419 when < 24 mo, P = 0.311 when > 24 mo). For dogs in groups 1 and 2 there was no significant linear relationship between time after surgery and OA score change (Figure 1). There were no significant findings between OA scores and independent variables, including weight, age, gender, and breed.

Figure 1.

For dogs in group 1, there was no significant linear relationship between time after surgery and OA score change. For dogs in group 2, there was no significant linear relationship between time after surgery and OA score change.

When comparing consistency amongst the 2 radiologists, the correlations were highly significant and the correlation coefficients ranged from 0.77 to 0.89. These results indicate that there were strong, linear relationships between the scores from the 2 radiologists. When the scores were evaluated by t-tests, the mean scores for the first radiologist were significantly greater than mean scores for the second radiologist. When comparing preoperative and follow-up scores, the average differences in scores between the first and second radiologists were 2.2 and 7.9, respectively, for the first operated stifle, and 3.3 and 7.1, respectively, for the second operated stifle.

Discussion

Results of the present study suggest osteoarthritis following TPLO surgery for correction of CCL rupture in dogs progresses over time, but with no evidence that the rate of progression changes. Progression of osteoarthritis following TPLO is not a new concept. Hurley et al (4) found a small but significant increase in mean degenerative joint disease (DJD) scores 8 wk after surgery, compared with preoperative scores in 373 stifles. Lazar et al (10) evaluated 66 dogs 1 y after TPLO or extracapsular stabilization to treat CCL rupture. They determined that OA scores increased following both procedures. Au et al (11) evaluated OA scores up to 2 y following TPLO or lateral fabellar suture stabilization. In all dogs, OA scores were significantly increased after 24 mo. Lineberger et al (9) studied how OA scores changed after TPLO with 2 different techniques of arthrotomy. OA scores were higher in the 72 stifles evaluated 7 to 38 mo following surgery. Boyd et al (12) evaluated dogs 12 to 33 mo following TPLO and found that 76% of dogs showed an increase in OA score. These studies dispute the claim that TPLO stops the progression of osteoarthritis; however, they do not establish the relationship between OA changes and time.

Rupture of the CCL creates joint instability in the injured stifle, which causes inflammation, with resultant osteoarthritis (8). Early in the process of osteoarthritis, there is an imbalance between peptides that promote healthy extracellular matrix, such as transforming growth factor beta and insulin-like growth factors, and substances that induce remodeling, such as cytokines interleukin-1 (IL-1) and tumor necrosis factor alpha, prostaglandin E2, and leukotriene B4 (16). It is reasonable to assume that inflammation begins to subside following stabilization of the joint, i.e., after TPLO surgery. It would also be reasonable to assume that the progression of osteoarthritis slows over time as inflammation becomes less significant. Based on these assumptions we expected to see a greater increase in OA scores in the early postoperative period. However, we found that OA score changes did not differ significantly between dogs in the early postoperative period compared with dogs that had longer follow-up times. This may suggest that either inflammation does not subside after the initial injury, or that other contributing factors continue to persist after TPLO surgery. These factors may be humoral, mechanical or surgical, and may have different roles in the progression of OA in patients.

We compared OA progression in dogs having surgery on 1 stifle to those having surgery on both stifles and found no significant difference. Concurrent meniscal damage may affect OA. Johnson et al (17) reported evidence of secondary osteoarthritis in dogs with intact cranial cruciate ligaments 16 wk after a caudal pole hemi-meniscectomy or complete medial meniscectomy. Lineberger et al (9) reported that dogs with CCL rupture receiving a limited caudal medial arthrotomy had significantly less progression of osteoarthritis compared with those receiving a medial parapatellar open arthrotomy. We were not able to make conclusions for all patients in this study regarding concurrent meniscal injury or surgery as this information was not always available. Another factor to consider is the chronicity of the cruciate injury. Chronic injuries may be associated with a higher preoperative OA score which may affect further progression compared with an acute ligament rupture.

Radiographic evidence of osteoarthritis may not relate to clinical function of the operated pelvic limb. Using force plate analysis as a measure of limb function, Gordon et al (14) evaluated dogs with clinical stifle OA that were not treated medically or surgically. They found that there was no significant relationship between platform data and radiographic OA score, and concluded that the presence of OA in the stifle found radiographically suggests stifle pathology, but does not correlate with clinical function. Au et al (11) evaluated thigh circumference and OA scores in dogs 6 mo and 24 mo after unilateral TPLO and found dogs returned to 98% symmetry at 6 mo and did not change significantly between 6 mo and 24 mo, despite significant increases in OA scores. These dogs underwent rigorous post-operative physiotherapy, which has been shown to improve outcome and function after surgery for CCL injury (18). Moeller et al (19) also demonstrated that thigh circumference returns to 98.5% of control (unoperated limb) 1 to 5 y after TPLO. Although OA scores were not evaluated we can assume, based on the studies previously mentioned, that osteoarthritis did progress in these dogs. Interestingly, time after surgery had no apparent affect on thigh circumference or stifle range of motion.

Typically, stifle radiographs are performed immediately after TPLO surgery, and then at recheck examinations between 8 to 12 wk after surgery. These are performed to monitor healing of the osteotomy, to look for changes in the implants that would indicate surgical failure, and to monitor the soft tissue structures of the stifle. Our study suggests that OA changes will progress past 12 wk, but further radiographic evaluation is usually not warranted in patients which are not clinically affected. Our study also shows that OA does not progress rapidly after surgery with a slow progression over time, or conversely, slowly after surgery and more rapidly with time. This may alter post-operative treatment options, including the administration of long-term NSAIDS, chondroprotective agents, joint-specific neutraceuticals, and other disease-modifying medications.

When evaluating OA scores based on signalment, we did not find any significant correlation between OA score and patient weight, age, or gender. One study showed an increase in DJD scores between dogs that received a 3.5-mm TPLO plate versus dogs that received 2.7-mm or 2.0-mm TPLO plates (4). They stated that larger dogs were affected by osteoarthritis more significantly than smaller dogs. Our study had a low number of small breed dogs. This may potentially skew our results when evaluating OA scores and patient size.

This study is retrospective and therefore was not standardized. There were several surgeons and residents performing surgery during the study period, and whether an arthrotomy was performed and the type of arthrotomy chosen was based on the surgeon’s preference. All follow-up radiographs were performed on awake patients, and although the views included in this study were deemed acceptable for OA evaluation, technique and positioning may have been suboptimal compared with preoperative views taken under anesthesia. The OA scoring system we used was modified from a previously devised system and is based on the assumption that osteophytosis is synonymous with osteoarthritis. Innes et al (15) found that osteophytosis was the most significant change over time, whereas subchondral sclerosis was not a reliable indicator of OA changes. Studies performed previously all use a different OA scale and so cannot be compared with the present study (8,10,14,15). Future studies may look to standardize a radiographic osteoarthritis scoring system to eliminate this as a variable. Additional studies may be conducted to compare OA changes in specific joint regions over time, and to identify factors other than time contributing to OA progression after TPLO surgery.

In conclusion, osteoarthritis did progress over time following TPLO surgery to correct CCL rupture in dogs, but the degree of progression was not time dependent. This finding may lead to alterations in patient monitoring and care following surgical stabilization of the stifle. CVJ