Metacarpophalangeal and Metatarsophalangeal Osteoarthritis in 49 Dogs
Although osteoarthritis (OA) is a common and debilitating condition in the canine patient, few data are available on OA of the metacarpophalangeal (MCP) and metatarsophalangeal (MTP) joints. Review of medical records of 49 dogs with a radiographic diagnosis of MCP or MTP OA presented over a 7-year period demonstrated that OA was an “incidental finding” for the majority of animals (n=35), while 14 dogs were identified as clinically lame as a result of MCP or MTP OA. Dogs that were clinically lame as a result of MCP or MTP OA were significantly more likely to have visible swelling over the affected digits. Five times as many dogs were diagnosed with MCP OA than with MTP OA, and the majority of dogs had radiographic changes on multiple digits. Review and scoring of radiographs (n=44 dogs) for six radiographic signs of OA followed by logistic regression analysis demonstrated that the two lateral digits of the front limb were significantly more likely to have osteophytosis and enthesophytosis than the two medial digits. Osteoarthritis of MCP and MTP joints has unique radiographic features that are not seen in the stifle, hip, shoulder, or elbow; these features can complicate accurate diagnosis, particularly differentiation from primary bone neoplasia. These data underscore the clinical relevance of this condition and provide useful information detailing which and how many digits are most commonly affected, potentially assisting discrimination between OA and neoplasia.
Introduction
Metatarsophalangeal (MTP) osteoarthritis (OA) and metacarpophalangeal (MCP) OA are common and well-described conditions in humans that are presumably secondary to either acute trauma or chronic, repetitive microtrauma.1–3 Accordingly, a plethora of medical and surgical treatments used in humans has been described to treat either acute injuries or chronic degenerative diseases of varying severity.2,4,5 Conversely, data on MCP and MTP OA in dogs are limited to just two reports in the German literature describing the treatment of four dogs with MCP OA by excision arthroplasty of the diseased joints.6,7 The absence of a thorough description of this syndrome is notable considering the prevalence and the impact that OA has on the quality of life for many dogs.8 Furthermore, the authors’ subjective impression has been that canine OA of these particular joints is not uncommon, has unique radiographic characteristics, and can be a cause of persistent and progressive lameness.
Purposes of this retrospective study were to review the medical records and radiographs of dogs that had a radiographic diagnosis of MCP or MTP OA to document signalment, clinical signs, radiographic signs, and digits involved for this population, and to determine if a correlation existed between clinical signs and radiographic signs.
Materials and Methods
Medical records at the Colorado State University Veterinary Teaching Hospital were reviewed for dogs that had a radiographic diagnosis of MCP or MTP OA between January 1, 2000 and January 1, 2007. Signalment, weight, presenting complaint, joints affected, presence/absence of soft-tissue swelling, and histological confirmation (if any) of a degenerative process were recorded for all dogs for which these data were available.
Cases were then divided into two groups for dogs that had radiographs available for review (n=44). The first group was composed of dogs with lameness attributed to MCP/MTP OA as a primary complaint (n=13). Group two consisted of dogs with MCP/MTP OA as an incidental finding secondary to some other pathology (n=31). All radiographs made of the MCP or MTP joints at any time (i.e., at the initial presentation and on follow-up visits) were then reviewed by a board-certified radiologist (Park) who was blinded to group assignment. Every MCP or MTP joint of the second through fifth digits was scored on the presence or absence of six radiographic signs: osteophytes, enthesophytes, joint space narrowing, subchondral bone sclerosis, fracture of a sesamoid bone, or an articular bone fragment.
Initially, osteophytes and enthesophytes were given a score on a four-point scale (i.e., 0 to 3), with a score of 0 representing no pathology and 3 representing the most severe pathology. Osteophytes are proliferations of bone on articular margins—most often at the periphery of articular cartilage. Enthesophytes are osseous spondylopathies of the origins/insertions of soft-tissue stabilizers of the joint, such as ligaments, muscle tendons, and the joint capsule.9 However, after initial analysis of 228 digits, only six digits were scored a 2 or 3 for osteophytes, and 39 scored a 2 or 3 for enthesophytes. These variables were therefore reassigned binary values, with anything originally assigned a score of ≥1 given a score of 1.
Data were then analyzed using logistic regression analysis, a with each of the six radiographic signs as the response variable for six different model runs. The explanatory variables for each model were the group assignment (primary complaint versus incidental finding), whether the joint was a MCP or MTP joint, and the digit examined; the digit was nested within the MCP/MTP variable. Since the data were binomial, the authors specified the response distribution as binomial and used the log default ilink function in GLIMMIX. 10 If multiple limbs for a single dog were radiographed, data for the different limbs from the same animal were included in the analysis. If repeated studies of the same limb were done for a given dog, only the most recent radiographic study for that limb was used. The animal, nested within its group assignment, was used as a random effect. A Tukey adjustment was made for all pair-wise comparisons to correct for multiple tests. A P value of <0.05 was considered significant. The difference in prevalence of soft-tissue swelling over the affected digits between dogs that were considered lame as a result of MCP or MTP OA and dogs in which OA of these joints was an incidental finding was evaluated post hoc using a Fisher’s exact test.
Results
Forty-nine dogs that presented to the Colorado State University Veterinary Teaching Hospital between January 1, 2000 and January 1, 2007 were identified with a radiographic diagnosis of MCP or MTP OA. For 35 dogs, MCP or MTP OA was an incidental finding. Twenty-three of the 35 dogs either had a mass (e.g., soft-tissue neoplasia, infection, or inflammation) or another orthopedic abnormality in the MCP/MTP region that prompted radiographic examination with subsequent documentation of OA, while 12 dogs had a primary neoplasia (usually osteosarcoma) elsewhere in the body. High rates of osteoblastic activity at the MCP or MTP joints were noted with bone scans of these 12 dogs, and follow-up radiographs revealed OA. Fourteen dogs had lameness attributable to MCP/MTP OA. In 13 of these 14 cases, owners reported acute on chronic episodes of lameness that varied in frequency and severity. Eleven of the 14 dogs had firm swelling over the affected joints. This is a significantly higher prevalence of swelling than that noted in the second (incidental finding) group (Fisher’s exact test, two-tail, P=0.01). Selected signalment and clinical signs for dogs in each group are presented in Table 1.
Histological evaluations of biopsies of affected bone were available for four dogs as described elsewhere;11 three of the four dogs were clinically lame as a result of MCP/MTP OA, and it was uncertain if the radiographic changes were consistent with OA or neoplasia. The fourth dog originally had a soft-tissue sarcoma removed from the affected paw, but at reexamination following surgery, radiographs provided evidence of bone changes. Concern of neoplasia prompted digit amputation (even though the dog was not lame), and the digit was submitted for histopathology. In all four dogs, periosteal reaction and/or reactive bone was present and found to be consistent with OA; no evidence of neoplasia was seen.
Radiographs were available for review for 44 dogs (228 digits). At least one of the six radiographic signs was identified on 104 digits. The average number of affected digits per dog was 2.4. Eighty-nine of the affected joints were MCP joints; 15 were MTP joints [Table 2]. Serial radiographs were available for five dogs that were clinically lame, and in every case progression of OA was confirmed by increasing scores of OA on radiograph reviews. Representative radiographs from three dogs, all of whom had histological confirmation of a degenerative process, are included [Figure 1].
The logistic regression analysis converged for those models that had osteophytes, enthesophytes, and joint space narrowing as the response variables. These models failed to detect a significant relationship between group allocation (primary complaint versus incidental finding) and whether radiographic pathology was present on the digit examined; the P value was >0.05 for the main effect of group assignment for the three models. However, the digit examined was a significant predictor of whether radiographic pathology was present. Specifically, the main effect of digit was not significant (P<0.07) for osteophytes, but a significant difference was found between the fifth digit and the third digit (P<0.03 after a Tukey adjustment for multiple pairwise comparisons). The fifth digit was significantly more likely to have osteophytes than the third digit. The main effect of digit was a significant predictor of the presence of enthesophytes (P<0.04), with the fourth digit significantly more likely to have enthesophytes than the second digit (P<0.03 after Tukey adjustment). The digit was not a significant predictor of whether joint space narrowing was noted (P>0.05).
The logistic regression analysis failed to converge for those models that evaluated bone sclerosis, articular fragments, and sesamoid fractures as the response variables. The authors suspect convergence failed because only six, eight, and 12 digits had identifiable subchondral bone sclerosis, articular fragments, or sesamoid fractures, respectively. The authors repeated these analyses, including just those dogs that were clinically lame, while using the same response variables but as a function of only the digit affected. The digit examined was not a significant predictor of radiographic pathology (P>0.05) for any of those three models.
Discussion
Review of the medical records at the authors’ institution identified 49 dogs over a 7-year period that had MCP or MTP OA. Although OA of these joints was considered an incidental finding for the majority (n=35) of the cases, 14 of the 49 dogs had lameness that was attributable to MCP/MTP OA. This is noteworthy, as it equates to two dogs per year that were lame due to a condition which has only been reported in one case series and one case report in the German literature.
The signalments of dogs with MCP/MTP OA were variable among both groups, with wide ranges in ages and body weights and no apparent gender or breed predispositions. The authors did not identify any statistically significant relationship between lameness as a result of MCP/MTP OA and the presence of six radiographic signs of OA. However, the authors did find that dogs in the primary complaint group were significantly more likely to have firm swelling over their affected digits [Table 1]. Swelling over the digits should raise suspicion that MCP/MTP OA may be causing lameness, and this should be an indication for making radiographs of these joints in a lame dog.
Approximately five times as many dogs had MCP involvement than MTP involvement, and six times as many MCP joints were affected than MTP joints. The authors’ results are consistent with results of a recent publication showing that palmar and plantar digital cysts (presumed to result from repetitive trauma) occurred on the front limbs almost four times as often as on the hind limbs.12 Metacarpophalangeal joints may be more frequently affected than MTP joints, because more weight is borne on the forelimbs—a finding that has been confirmed with force plate analysis in Labrador retrievers and greyhounds.13
Although radiographic signs of OA of MCP and MTP joints can be similar to signs of OA seen elsewhere in the body, some radiographic findings are unique to these joints and make radiographic diagnosis problematic. Most notably, a profound periosteal reaction extending as far as 33% the length of the metacarpals or metatarsals was noted in several study dogs [Figure 2]. This proliferative response could be mistaken for primary bone neoplasia. However, osteophytosis and enthesopathy observed in the dogs with OA in this study were usually identifiable on both sides of the joint, they were smooth, and they did not completely destroy the cortex. Conversely, primary bone neoplasia of the metacarpal and metatarsal bones rarely crosses the joint space and can cause complete cortical lysis extending from the endosteal surface.11,14
Imaging multiple limbs and evaluating the number of digits with osseous changes may also be helpful in differentiating OA from neoplasia, as primary osteosarcoma of the metacarpals or metatarsals arises from a single metacarpal or metatarsal bone.14 Of 44 dogs in this study for which the authors reviewed radiographs, 27 had evidence of OA on multiple digits; an average of 2.4 digits were affected per animal [Table 2]. Furthermore, all 17 dogs that had only one digit affected had radiographs of only one limb, while all 11 dogs for which radiographs of multiple limbs were available had radiographic signs of OA on more than one digit. Possibly a greater number of digits with MCP/MTP OA would have been identified had radiographs of more limbs been available for each animal. Films from multiple limbs may also facilitate differentiation from metastatic bone tumors that may affect multiple bones in a single paw but would be less likely to create osseous changes centered at the MCP/MTP joints on multiple paws.
Histological evaluations of bone biopsies are recommended for cases in which differentiating OA from neoplasia remains problematic even after radiographs of multiple limbs have been made.11 Histological evaluations of bone biopsies were performed for four of the dogs in this study, and the results were consistent with degenerative changes rather than neoplasia in all four dogs (including one case in which the digit was amputated based upon suspicion of neoplasia). Making the appropriate diagnosis before digit amputation is important, because amputation of a digit exacerbates degenerative changes in the adjacent digits.7
The two lateral digits (fourth and fifth) were significantly more likely to be affected by osteophytosis and enthesophytosis, respectively, than the two medial digits. Identifying osteophytes and enthesophytes was relatively easy, but differentiating between the two was difficult in this study. Therefore, it may be better to generalize that the outer two digits are more likely to be affected by osteoarthritis than be predisposed to osteophytes or enthesophytes specifically. Unfortunately, the utility of these findings in assisting diagnosis of OA is dubious, as OA was identified on all digits and was not completely restricted to the fourth and fifth digits.
As with many disease processes, understanding pathogenesis would be helpful in guiding treatment. One report in the German literature hypothesizes that repetitive microtrauma is the most likely cause; this hypothesis stems from the observation that working dogs, which are repeatedly placing strain on these joints, are most frequently affected.7 Support for this premise can be found in the human literature; repetitive microtrauma is the most frequently purported cause of proliferative OA of both the first MCP and MTP joints in humans. The term “Gamekeeper’s thumb” was coined to describe degeneration of the first MCP, because it was first identified in Scottish gamekeepers in the 1950s. These individuals suffered repetitive microtrauma from gripping the heads of injured rabbits between the thumb and index finger.2 Likewise, Hallux rigidus (or Hallux limitus) is an extremely common condition that involves degenerative arthritic changes of the first MTP joint; it is estimated to affect one in 40 people over the age of 50, and the most frequently cited cause is repetitive microtrauma.1 Alternatively, acute hyperextension injury could predispose to joint laxity and subsequent OA development. Skier’s thumb (i.e., forceful abduction of the thumb by the ski pole during a fall) and “turf toe” (i.e., hyperextension of the big toe most frequently recognized in football players competing on artificial turf) are acute hyperextension injuries that have been shown to predispose to chronic OA of the first MCP and MTP joints, respectively.2,15
The data of the authors’ study are not entirely consistent with traumatic hyperextension or repetitive microtrauma as the cause of MCP/MTP OA. If acute hyperextension injuries were the cause of MCP/MTPOA, the authors would expect to find single digits most commonly affected, as with humans having skier’s thumb and turf toe. Conversely, if repetitive microtrauma were the most common cause, the authors would expect to find the third and fourth digits most commonly affected since they are the two primary weight-bearing digits.13 However, force plate analysis has shown that the fifth digit also bears a substantial amount of load.13 Additionally, interdigital cysts (considered to be the result of repeated trauma as opposed to an acute hyperextension injury) have been shown to occur most frequently between the fourth and fifth digits, far less so between the third and fourth digits, and even less frequently between the second and third digits. The propensity for two different disease processes, suspected to result from some form of trauma (i.e., interdigital cysts and OA), to occur on the lateral digits suggests that the lateral two digits may be more prone to traumatic injuries than the medial two digits. Considering that interdigital cysts are not the result of acute hyperextension injuries, these data may be more consistent with the lateral two digits being more prone to repetitive microtrauma as opposed to acute hyperextension injuries.
Just as the exact etiology of this condition is unclear, so is the ideal treatment. Jauernig et al (1999) briefly discussed the inefficacy of medical management and digit amputation. They described treatment of three dogs with excisional joint arthroplasty; two dogs had excellent outcomes, and a satisfactory result was reported in the single dog that had excision of the MCP joints from the third and fourth digits.7 Excision arthroplasty was also reported as a successful treatment in one other case of MCP OA and in two cases with pathology of the interphalangeal joints.6,16 Although arthrodesis has not been described as a treatment for chronic OA of the MCP/MTP joints, it has been reported as providing a good outcome in two of three animals with osteomyelitis and joint degeneration following trauma and in one animal with immune-mediated arthritis.17,18 Ultimately, more information is needed to evaluate the results of medical and surgical management. Furthermore, investigation is warranted into alternative therapies, such as use of orthotics or joint implant arthroplasty, which have frequently been used in correlating human conditions.
Conclusion
Metacarpophalangeal or MTP OA is usually an incidental finding but should be considered a potential cause of lameness, particularly when firm swelling is noted over these joints. Osteoarthritis of the MCP joints was more commonly diagnosed than MTP OA. The majority of dogs in this study had radiographic evidence of OA on multiple digits; this may help to differentiate the disease from metacarpal or metatarsal osteosarcoma. For those dogs that were followed over time, repeat radiographs always showed progression of degenerative changes. The two lateral digits were more frequently affected than the two medial digits; the clinical significance of this remains unclear.
Acknowledgments
The authors thank Dr. zumBrunnen for performing the statistical analyses. A. Fields assisted with data entry.
PROC GLIMMIX, SAS version 9.1; SAS, Cary, NC 27513.
Citation: Journal of the American Animal Hospital Association 45, 3; 10.5326/0450112
Citation: Journal of the American Animal Hospital Association 45, 3; 10.5326/0450112
Radiographs from three dogs that were lame and had histological confirmation of metacarpophalangeal or metatarsophalangeal osteoarthritis. The third and fourth digits are affected in dog A; the second, fourth, and fifth digits are affected in dog B; and the second digit is affected in dog C.
Radiographs highlighting a profound periosteal reaction extending proximally along the fourth metacarpal and which could be confused with a primary bone neoplasia. Note the enthesopathy on the proximal aspect of the first phalanx of the same digit.
